Prolongation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "75fa5d6f-e543-48f1-92d5-f0740ace8da0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import os\n",
    "import librosa\n",
    "import librosa.display\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import sklearn.preprocessing\n",
    "from scipy.signal import lfilter\n",
    "from joblib import Parallel, delayed\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.metrics import accuracy_score, classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "65581d99-dbf0-4da8-8c50-52495a5cc0f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(r\"C:\\Users\\sakshi\\Desktop\\SMLProject\\bin_labeled_w_audio_path.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "96f226f2-970d-4cee-b983-ea4cd9456cd4",
   "metadata": {},
   "outputs": [],
   "source": [
    "clip_path = r\"C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtraction\\extracted_clips\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f753af9a-5074-4d95-93d9-b8e17b6ec4f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Stutter\n",
      "0    60.677123\n",
      "1    39.322877\n",
      "Name: count, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "value_counts = df['Stutter'].value_counts()\n",
    "total_rows = df.shape[0]\n",
    "\n",
    "percentage_per_class = (value_counts / total_rows) * 100\n",
    "\n",
    "print(percentage_per_class)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "96550543-7eef-4257-9233-71b923ba9cc5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_mfcc(audio_file, num_mfcc=40):\n",
    "    audio, sr = librosa.load(audio_file, sr=None)\n",
    "    mfccs = librosa.feature.mfcc(y=audio, sr=sr, n_mfcc=num_mfcc)\n",
    "    mfccs_transposed = mfccs.T  # Transpose the MFCC matrix\n",
    "    mfccs_flat = mfccs_transposed.flatten()  # Flatten the transposed MFCC matrix\n",
    "    return mfccs_flat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c123d444-9649-488c-ac76-5e333781ca38",
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_zcr(audio_file, frame_length=2048, hop_length=512):\n",
    "    audio, sr = librosa.load(audio_file, sr=None)\n",
    "    zcr = librosa.feature.zero_crossing_rate(audio, frame_length=frame_length, hop_length=hop_length)\n",
    "    return zcr.flatten()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7b7f8a4b-832e-4828-96fd-6825d38d87f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_jitter(audio_file, frame_length=2048, hop_length=512):\n",
    "    y, sr = librosa.load(audio_file, sr=None)\n",
    "    f0, voiced_flag, voiced_probs = librosa.pyin(y, fmin=75, fmax=600)\n",
    "    periods = 1 / f0[voiced_flag]\n",
    "    periods[np.isnan(periods)] = 0\n",
    "    period_diffs = np.diff(periods)\n",
    "    jitter = np.mean(np.abs(period_diffs)) * 100 \n",
    "    return jitter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f76cefbd-cb92-4d25-84d3-feb69ff7fd7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_shimmer(audio_file, frame_length=2048, hop_length=512):\n",
    "    y, sr = librosa.load(audio_file, sr=None)\n",
    "    f0, voiced_flag, _ = librosa.pyin(y, fmin=75, fmax=600)\n",
    "    rms = librosa.feature.rms(y=y, frame_length=frame_length, hop_length=hop_length)[0]\n",
    "    voiced_rms = rms[voiced_flag]\n",
    "    voiced_rms[np.isnan(voiced_rms)] = 0\n",
    "    shimmer = np.mean(np.abs(np.diff(voiced_rms))) / np.mean(voiced_rms) * 100\n",
    "    return shimmer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "54a955e8-ad68-428e-9f13-00b637cfe19d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def merge_features(mfcc, zcr, jitter, shimmer):\n",
    "    return np.hstack((mfcc, zcr, jitter, shimmer))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3762825a-3adb-4d8f-a961-1c35281392f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "audio_files = df['audio_path'] \n",
    "labels = df['Prolongation']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "17a5dc77-a920-40d6-890e-29226c4d04da",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_prolongation = df[['audio_path', 'Prolongation']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "2729371f-f954-49ff-ba12-44b64797b2cf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>audio_path</th>\n",
       "      <th>Prolongation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19696</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19697</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19698</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19699</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19700</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19701 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              audio_path  Prolongation\n",
       "0      C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "1      C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "2      C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "3      C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "4      C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "...                                                  ...           ...\n",
       "19696  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "19697  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "19698  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "19699  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "19700  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...             0\n",
       "\n",
       "[19701 rows x 2 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_prolongation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "7ff9cec9-75db-4404-b717-8bc192b3b574",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Prolongation\n",
       "0    17636\n",
       "1     2065\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Prolongation'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "19feca2f-a112-4063-818f-207ea944b251",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_prolongation_1 = df_prolongation[df_prolongation['Prolongation']==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "bc2274ff-5093-4610-a2a7-b948261be911",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_prolongation_0 = df[(df['Prolongation']==0) & \n",
    "                        (df['Block']==0) & \n",
    "                        (df['Interjection']==0) & \n",
    "                        (df['WordRep']==0) & \n",
    "                        (df['SoundRep']==0) &\n",
    "                        (df['NoStutteredWords']==1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "7b508592-f68c-4875-925c-f2565fc2de9a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2065, 2)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_prolongation_1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a4693422-971b-4b04-a7f0-50c684f66088",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_prolongation_0_sampled = df_prolongation_0.sample(df_prolongation_1.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b8f16881-7b46-4202-9950-c74b297ac9b5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2065, 20)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_prolongation_0_sampled.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "64753d0b-0fa7-43eb-b1c1-a94c01edc210",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_prolongation_data = pd.concat([df_prolongation_0_sampled, df_prolongation_1], ignore_index=True, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "fcccbaf6-ced7-4c29-85b6-16ff1303db81",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4130, 20)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_prolongation_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "55ff8824-0492-41f2-a10d-b13342dea20a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0.2</th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>Show</th>\n",
       "      <th>EpId</th>\n",
       "      <th>ClipId</th>\n",
       "      <th>Start</th>\n",
       "      <th>Stop</th>\n",
       "      <th>Prolongation</th>\n",
       "      <th>Block</th>\n",
       "      <th>SoundRep</th>\n",
       "      <th>WordRep</th>\n",
       "      <th>Interjection</th>\n",
       "      <th>NoStutteredWords</th>\n",
       "      <th>NaturalPause</th>\n",
       "      <th>NoSpeech</th>\n",
       "      <th>audio_path</th>\n",
       "      <th>Label</th>\n",
       "      <th>audio_length</th>\n",
       "      <th>Stutter</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18298.0</td>\n",
       "      <td>20335.0</td>\n",
       "      <td>26656.0</td>\n",
       "      <td>WomenWhoStutter</td>\n",
       "      <td>92.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>34187040.0</td>\n",
       "      <td>34235040.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48000.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>455.0</td>\n",
       "      <td>546.0</td>\n",
       "      <td>546.0</td>\n",
       "      <td>HeStutters</td>\n",
       "      <td>3.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>20033760.0</td>\n",
       "      <td>20081760.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48000.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9337.0</td>\n",
       "      <td>10320.0</td>\n",
       "      <td>12628.0</td>\n",
       "      <td>StutterTalk</td>\n",
       "      <td>43.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>61300480.0</td>\n",
       "      <td>61348480.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48000.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11089.0</td>\n",
       "      <td>12219.0</td>\n",
       "      <td>14527.0</td>\n",
       "      <td>StutterTalk</td>\n",
       "      <td>72.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>10928640.0</td>\n",
       "      <td>10976640.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48000.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12295.0</td>\n",
       "      <td>13623.0</td>\n",
       "      <td>19944.0</td>\n",
       "      <td>WomenWhoStutter</td>\n",
       "      <td>6.0</td>\n",
       "      <td>119.0</td>\n",
       "      <td>2137440.0</td>\n",
       "      <td>2185440.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48000.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0.2  Unnamed: 0.1  Unnamed: 0             Show  EpId  ClipId  \\\n",
       "0       18298.0       20335.0     26656.0  WomenWhoStutter  92.0    28.0   \n",
       "1         455.0         546.0       546.0       HeStutters   3.0    17.0   \n",
       "2        9337.0       10320.0     12628.0      StutterTalk  43.0    11.0   \n",
       "3       11089.0       12219.0     14527.0      StutterTalk  72.0    87.0   \n",
       "4       12295.0       13623.0     19944.0  WomenWhoStutter   6.0   119.0   \n",
       "\n",
       "        Start        Stop  Prolongation  Block  SoundRep  WordRep  \\\n",
       "0  34187040.0  34235040.0             0    0.0       0.0      0.0   \n",
       "1  20033760.0  20081760.0             0    0.0       0.0      0.0   \n",
       "2  61300480.0  61348480.0             0    0.0       0.0      0.0   \n",
       "3  10928640.0  10976640.0             0    0.0       0.0      0.0   \n",
       "4   2137440.0   2185440.0             0    0.0       0.0      0.0   \n",
       "\n",
       "   Interjection  NoStutteredWords  NaturalPause  NoSpeech  \\\n",
       "0           0.0               1.0           0.0       0.0   \n",
       "1           0.0               1.0           0.0       0.0   \n",
       "2           0.0               1.0           0.0       0.0   \n",
       "3           0.0               1.0           0.0       0.0   \n",
       "4           0.0               1.0           0.0       0.0   \n",
       "\n",
       "                                          audio_path  Label  audio_length  \\\n",
       "0  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...    NaN       48000.0   \n",
       "1  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...    NaN       48000.0   \n",
       "2  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...    NaN       48000.0   \n",
       "3  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...    NaN       48000.0   \n",
       "4  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...    NaN       48000.0   \n",
       "\n",
       "   Stutter  \n",
       "0      0.0  \n",
       "1      0.0  \n",
       "2      0.0  \n",
       "3      0.0  \n",
       "4      0.0  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_prolongation_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "86cb8b2d-7035-457c-9f24-8454047816d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_prolongation_data.drop(['Unnamed: 0.2','Unnamed: 0.1','Unnamed: 0'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "0b3eeada-ce78-494c-87d1-2b275e5380a0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prolongation\n",
      "0    50.0\n",
      "1    50.0\n",
      "Name: count, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "value_counts = df_prolongation_data['Prolongation'].value_counts()\n",
    "total_rows = df_prolongation_data.shape[0]\n",
    "\n",
    "percentage_per_class = (value_counts / total_rows) * 100\n",
    "\n",
    "print(percentage_per_class)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "be1a9b7b-1544-40d3-9c8e-3c25dead677b",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.array(df_prolongation_data['audio_path'])\n",
    "y = np.array(df_prolongation_data['Prolongation'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "503c8b42-b9de-41fc-ac80-b438736d00e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=40, stratify = y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "4cf6ca75-dc98-4a43-b86b-773c9cdce806",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_temp_df = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "3b1e94a1-fc79-4d67-ada4-3949d009b07f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract MFCC features for training set\n",
    "train_temp_df['mfcc'] = Parallel(n_jobs=-1)(delayed(extract_mfcc)(audio_file) for audio_file in X_train)\n",
    "\n",
    "# Extract ZCR features for training set\n",
    "train_temp_df['zcr'] = Parallel(n_jobs=-1)(delayed(extract_zcr)(audio_file) for audio_file in X_train)\n",
    "\n",
    "# Extract Jitter features for training set\n",
    "train_temp_df['jitter'] = Parallel(n_jobs=-1)(delayed(calculate_jitter)(audio_file) for audio_file in X_train)\n",
    "\n",
    "# Replace NaN values in Jitter column with zero for training set\n",
    "train_temp_df['jitter'] = train_temp_df['jitter'].fillna(0)\n",
    "\n",
    "# Extract Shimmer features for training set\n",
    "train_temp_df['shimmer'] = Parallel(n_jobs=-1)(delayed(calculate_shimmer)(audio_file) for audio_file in X_train)\n",
    "\n",
    "# Replace NaN values in Shimmer column with zero for training set\n",
    "train_temp_df['shimmer'] = train_temp_df['shimmer'].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "1d6ee7bc-899a-4dc6-ae0f-99d9865a0cec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "mfcc       0\n",
       "zcr        0\n",
       "jitter     0\n",
       "shimmer    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_temp_df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "c42c6e40-b8c9-45cd-b31e-c18dba23411d",
   "metadata": {},
   "outputs": [],
   "source": [
    "mfcc_features_train = np.array(train_temp_df['mfcc'])\n",
    "zcr_features_train = np.array(train_temp_df['zcr'])\n",
    "jitter_features_train = np.array(train_temp_df['jitter'])\n",
    "shimmer_features_train = np.array(train_temp_df['shimmer'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "98b2cd80-38dc-4553-b887-a053256460b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "merged_features_train = [merge_features(mfcc, zcr, jitter, shimmer) for mfcc, zcr, jitter, shimmer in zip(mfcc_features_train, zcr_features_train, jitter_features_train, shimmer_features_train)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "29008b82-fc03-4c77-99b8-3edfd1782322",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([-5.77527893e+02, -1.33375473e+01,  6.25851536e+00, ...,\n",
       "         3.12500000e-02,  3.18565343e-02,  1.56339750e+01]),\n",
       " array([-1.73955841e+02,  3.43662338e+01,  6.11304712e+00, ...,\n",
       "         8.49609375e-02,  1.10148932e-02,  1.44385085e+01]),\n",
       " array([-2.50678818e+02,  1.10124664e+02,  3.26077957e+01, ...,\n",
       "         6.34765625e-02,  1.29674647e-02,  1.23009160e+01]),\n",
       " array([-6.09596313e+02,  3.80950928e+01,  2.19912109e+01, ...,\n",
       "         1.41113281e-01,  1.88549504e-02,  2.33311996e+01]),\n",
       " array([-4.74027649e+02,  7.02176743e+01,  4.53257179e+00, ...,\n",
       "         2.68554688e-01,  1.50230014e-02,  3.21851760e+01]),\n",
       " array([-1.54225479e+02,  1.12640793e+02,  1.24409370e+01, ...,\n",
       "         1.57226562e-01,  1.41888092e-02,  1.89927235e+01]),\n",
       " array([-2.52046188e+02,  7.26609192e+01, -1.71191940e+01, ...,\n",
       "         7.32421875e-02,  7.54464201e-03,  1.49818450e+01]),\n",
       " array([-6.32839050e+02,  4.31729507e+01,  2.33651085e+01, ...,\n",
       "         8.78906250e-02,  1.21332607e-02,  1.72803700e+01]),\n",
       " array([-3.06121155e+02,  9.29429779e+01,  2.79374828e+01, ...,\n",
       "         1.55761719e-01,  7.24219815e-03,  1.21890269e+01]),\n",
       " array([-2.01847656e+02,  1.81851639e+02, -2.33955002e+01, ...,\n",
       "         3.97460938e-01,  1.07297004e-02,  2.03102261e+01]),\n",
       " array([-5.20065796e+02,  3.38941841e+01,  1.46299114e+01, ...,\n",
       "         1.09863281e-01,  5.87794246e-03,  1.39476731e+01]),\n",
       " array([-1.74842545e+02,  1.12446701e+02, -7.30375443e+01, ...,\n",
       "         1.28417969e-01,  1.90346588e-02,  2.10280269e+01]),\n",
       " array([-2.05227432e+02, -5.48833008e+01,  2.43646107e+01, ...,\n",
       "         1.77734375e-01,  3.73935920e-02,  1.93066314e+01]),\n",
       " array([-3.09585571e+02,  7.36506958e+01,  3.35389197e-01, ...,\n",
       "         1.95312500e-02,  1.60920579e-02,  1.19242027e+01]),\n",
       " array([-7.75364304e+01,  1.22558784e+02, -6.11685562e+01, ...,\n",
       "         7.66601562e-02,  4.26254402e-02,  1.83677897e+01]),\n",
       " array([-1.26907211e+02,  9.81696625e+01,  3.46738091e+01, ...,\n",
       "         1.43554688e-01,  7.14164260e-03,  1.84631467e+01]),\n",
       " array([-5.61069702e+02,  2.09662437e+01,  1.46267252e+01, ...,\n",
       "         9.13085938e-02,  1.54184897e-02,  1.68351814e+01]),\n",
       " array([-7.13441528e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         4.58984375e-02,  1.29767585e-02,  1.37411967e+01]),\n",
       " array([-5.33256836e+02,  3.21329880e+00,  3.10105848e+00, ...,\n",
       "         5.12695312e-02,  1.16454695e-02,  1.30657420e+01]),\n",
       " array([-5.77884705e+02,  1.82549458e+01,  6.92542839e+00, ...,\n",
       "         4.68750000e-02,  1.03990258e-02,  1.53743029e+01]),\n",
       " array([-1.31025330e+02,  1.73494110e+02,  1.30232382e+00, ...,\n",
       "         1.88476562e-01,  2.03443111e-02,  1.45205632e+01]),\n",
       " array([-1.22001831e+02, -6.36739960e+01, -3.51673622e+01, ...,\n",
       "         2.57324219e-01,  1.96758926e-02,  1.82729617e+01]),\n",
       " array([-4.70160828e+02,  8.37168427e+01,  3.60164871e+01, ...,\n",
       "         4.29687500e-02,  1.89939360e-02,  1.89392120e+01]),\n",
       " array([-2.99501282e+02, -1.21532806e+02,  4.59568558e+01, ...,\n",
       "         2.44140625e-01,  2.08952034e-02,  2.48501465e+01]),\n",
       " array([-4.95110809e+02,  4.29115582e+00, -2.12772484e+01, ...,\n",
       "         2.09960938e-02,  1.73651649e-02,  1.51681766e+01]),\n",
       " array([-1.56196091e+02,  1.18006447e+02,  1.62839832e+01, ...,\n",
       "         3.02734375e-02,  1.04943848e-02,  1.35717005e+01]),\n",
       " array([-1.09305458e+02,  1.25111870e+02, -6.70908813e+01, ...,\n",
       "         2.83203125e-02,  1.00094841e-02,  1.08370505e+01]),\n",
       " array([-1.65082352e+02,  1.29412994e+02, -3.15416737e+01, ...,\n",
       "         6.54296875e-02,  1.81711295e-02,  1.30914301e+01]),\n",
       " array([-2.80564545e+02,  7.57663574e+01, -8.57098007e+00, ...,\n",
       "         8.44726562e-02,  1.76262878e-02,  1.73791230e+01]),\n",
       " array([-3.94219604e+02,  1.77064575e+02,  3.04562092e+01, ...,\n",
       "         5.37109375e-02,  2.23269006e-02,  1.97016567e+01]),\n",
       " array([-2.00111740e+02,  1.31369415e+02, -2.05318871e+01, ...,\n",
       "         1.19140625e-01,  2.05658750e-02,  1.92132398e+01]),\n",
       " array([-4.92207825e+02,  5.30532074e+01,  2.61625881e+01, ...,\n",
       "         5.17578125e-02,  1.42276192e-02,  1.58374712e+01]),\n",
       " array([-1.89457123e+02,  1.18003067e+02,  8.37643051e+00, ...,\n",
       "         0.00000000e+00,  3.13947185e-02,  1.84922099e+01]),\n",
       " array([-2.36804764e+02, -3.99046936e+01,  2.16908436e+01, ...,\n",
       "         6.83593750e-02,  5.86397223e-02,  4.03092563e+01]),\n",
       " array([-3.42778046e+02,  1.67365219e+02,  2.97785378e+01, ...,\n",
       "         4.78515625e-02,  1.46388501e-02,  1.44544587e+01]),\n",
       " array([-5.36717346e+02, -9.53841686e+00, -5.69399071e+00, ...,\n",
       "         2.29003906e-01,  4.32800844e-03,  2.44464576e+01]),\n",
       " array([-2.16435699e+02,  6.22904816e+01, -5.70711441e+01, ...,\n",
       "         4.68750000e-02,  1.99746705e-02,  2.08908007e+01]),\n",
       " array([-2.09613571e+02,  1.94519775e+02,  3.64479899e-01, ...,\n",
       "         9.47265625e-02,  7.50518036e-03,  9.68119651e+00]),\n",
       " array([-5.72791626e+02,  5.93609581e+01,  2.36839561e+01, ...,\n",
       "         1.85546875e-02,  2.58376753e-02,  2.14077383e+01]),\n",
       " array([-1.49571991e+02,  1.27481453e+02, -8.24941921e+00, ...,\n",
       "         1.75781250e-02,  2.84727011e-02,  2.44285509e+01]),\n",
       " array([-1.67357422e+02,  8.70445251e+01,  1.19028435e+01, ...,\n",
       "         2.86132812e-01,  8.68198423e-03,  1.54026225e+01]),\n",
       " array([-2.89340668e+02,  1.32434418e+02,  3.66891441e+01, ...,\n",
       "         7.81250000e-02,  2.01770320e-02,  1.88053727e+01]),\n",
       " array([-5.16018311e+02,  2.56026516e+01,  2.32637711e+01, ...,\n",
       "         3.56445312e-02,  1.07728044e-02,  2.70683169e+01]),\n",
       " array([-5.33236694e+02,  2.78619111e-01,  2.62080193e-01, ...,\n",
       "         2.34863281e-01,  8.34015566e-03,  2.00069115e+01]),\n",
       " array([-9.41066589e+01,  2.64888000e+01, -4.38996048e+01, ...,\n",
       "         1.70898438e-02,  7.05496683e-03,  1.92378610e+01]),\n",
       " array([-4.80865662e+02,  5.01603661e+01,  3.61355286e+01, ...,\n",
       "         2.98828125e-01,  8.81379353e-03,  1.44835964e+01]),\n",
       " array([-5.57954590e+02,  2.87511597e+01,  1.37558575e+01, ...,\n",
       "         4.79003906e-01,  6.70612151e-03,  1.48002684e+01]),\n",
       " array([-4.77880005e+02,  1.40891809e+01,  1.54935169e+01, ...,\n",
       "         1.08398438e-01,  9.05319521e-03,  1.93149477e+01]),\n",
       " array([-2.78238586e+02,  1.42095154e+02,  2.96275520e+01, ...,\n",
       "         3.86230469e-01,  1.08890285e-02,  1.27141714e+01]),\n",
       " array([-5.66161438e+02,  6.31596446e-01,  6.02826059e-01, ...,\n",
       "         2.08007812e-01,  7.35662990e-03,  2.89224535e+01]),\n",
       " array([-5.55563904e+02,  6.42664108e+01,  4.07969513e+01, ...,\n",
       "         1.23535156e-01,  1.39954242e-02,  3.38217318e+01]),\n",
       " array([-6.03978027e+02,  3.11921463e+01, -1.44187794e+01, ...,\n",
       "         8.74023438e-02,  6.54953034e-03,  1.21648125e+01]),\n",
       " array([-2.19420105e+02,  1.72011963e+02,  2.14692993e+01, ...,\n",
       "         1.37207031e-01,  1.97838681e-02,  2.07705468e+01]),\n",
       " array([-1.64885788e+02,  1.30467163e+02,  6.94975185e+00, ...,\n",
       "         1.60156250e-01,  1.44744522e-02,  1.65405363e+01]),\n",
       " array([-6.80038513e+02,  2.07300034e+01,  1.74852028e+01, ...,\n",
       "         1.23046875e-01,  2.04601009e-02,  1.02009058e+01]),\n",
       " array([-1.92828781e+02,  1.25238647e+02,  3.09565485e-01, ...,\n",
       "         4.19921875e-02,  1.17705064e-02,  2.30395302e+01]),\n",
       " array([-4.96144775e+02,  9.12406006e+01,  3.44658432e+01, ...,\n",
       "         4.39453125e-02,  3.69000011e-02,  3.24499816e+01]),\n",
       " array([-2.20109085e+02,  1.34535904e+02,  4.73458481e+01, ...,\n",
       "         3.40332031e-01,  3.70222126e-03,  1.40463024e+01]),\n",
       " array([-5.14014648e+02,  1.80198121e+00,  7.62700176e+00, ...,\n",
       "         1.75781250e-01,  5.28638787e-03,  1.42005563e+01]),\n",
       " array([-1.98017090e+02,  1.78721054e+02, -4.93952637e+01, ...,\n",
       "         3.27148438e-02,  1.16848622e-02,  1.89985022e+01]),\n",
       " array([-4.77241669e+02,  3.48529587e+01,  3.18253536e+01, ...,\n",
       "         5.90820312e-02,  1.52130907e-02,  2.45757073e+01]),\n",
       " array([-5.93047913e+02,  1.58076935e+01,  8.21527481e+00, ...,\n",
       "         1.20605469e-01,  1.54207397e-02,  2.81115800e+01]),\n",
       " array([-3.41618408e+02,  4.72873535e+01,  2.39232655e+01, ...,\n",
       "         9.52148438e-02,  6.04772216e-03,  1.55502051e+01]),\n",
       " array([-5.93832825e+02,  1.20022736e+01,  9.43837929e+00, ...,\n",
       "         4.83398438e-02,  3.41424132e-02,  1.23210706e+01]),\n",
       " array([-3.49347137e+02,  6.91076660e+01,  5.37916803e+00, ...,\n",
       "         1.04492188e-01,  2.79231523e-02,  1.49609759e+01]),\n",
       " array([-3.76813629e+02,  3.37120628e+01,  1.01056343e+02, ...,\n",
       "         4.10156250e-02,  1.51471707e-02,  1.64557457e+01]),\n",
       " array([-4.95775452e+02,  6.41422424e+01,  4.85812302e+01, ...,\n",
       "         1.22070312e-02,  8.47364281e-03,  1.42558575e+01]),\n",
       " array([-5.12787903e+02,  3.09194298e+01,  1.24648514e+01, ...,\n",
       "         1.07421875e-01,  1.17249679e-02,  1.89573094e+01]),\n",
       " array([-2.94347534e+02,  9.62180328e+01,  2.16324539e+01, ...,\n",
       "         3.56445312e-02,  1.63335880e-02,  1.25646815e+01]),\n",
       " array([-6.22227112e+02,  4.60189896e+01,  3.13092728e+01, ...,\n",
       "         5.27343750e-02,  2.40652072e-02,  2.25504547e+01]),\n",
       " array([-4.87036865e+02,  3.91372833e+01,  1.78981094e+01, ...,\n",
       "         4.34570312e-02,  1.14266117e-02,  1.80505082e+01]),\n",
       " array([-6.22919861e+02,  1.11171236e+01,  6.02025890e+00, ...,\n",
       "         1.53320312e-01,  1.01330476e-02,  2.24805564e+01]),\n",
       " array([-1.07038780e+02,  1.18707222e+02, -5.78242588e+00, ...,\n",
       "         1.48925781e-01,  4.85697058e-03,  1.69773400e+01]),\n",
       " array([-5.87540527e+02,  4.18074608e+00,  2.47301602e+00, ...,\n",
       "         6.44531250e-02,  1.32673318e-02,  1.77681699e+01]),\n",
       " array([-2.95062927e+02,  1.58219208e+02, -1.54774208e+01, ...,\n",
       "         0.00000000e+00,  1.55922387e-02,  2.03986660e+01]),\n",
       " array([-5.12826660e+02,  6.69067078e+01, -3.22002554e+00, ...,\n",
       "         1.30371094e-01,  1.39046029e-02,  1.75230205e+01]),\n",
       " array([-5.29386963e+02,  2.49759254e+01, -2.32632542e+01, ...,\n",
       "         6.93359375e-02,  8.70497153e-03,  1.99331090e+01]),\n",
       " array([-1.32314941e+02,  1.32931488e+02, -6.41642456e+01, ...,\n",
       "         3.66210938e-02,  1.13869097e-02,  1.04603007e+01]),\n",
       " array([-2.78279236e+02, -1.13640457e+02,  1.88303292e-01, ...,\n",
       "         1.55761719e-01,  4.40860801e-02,  1.90158442e+01]),\n",
       " array([-6.71121338e+02,  9.65312195e+00,  6.49896765e+00, ...,\n",
       "         9.76562500e-02,  1.90632899e-02,  2.02256650e+01]),\n",
       " array([-1.89642960e+02,  2.03023407e+02, -9.76970596e+01, ...,\n",
       "         6.59179688e-02,  1.22825086e-02,  1.50732651e+01]),\n",
       " array([-3.99505585e+02,  1.13018082e+02, -4.09068823e+00, ...,\n",
       "         1.11816406e-01,  3.54312613e-03,  1.47189692e+01]),\n",
       " array([-4.60481110e+02,  5.95262299e+01, -3.01927781e+00, ...,\n",
       "         6.29882812e-02,  1.79614929e-02,  1.76014945e+01]),\n",
       " array([-4.72676178e+02,  8.82186356e+01,  1.86742935e+01, ...,\n",
       "         1.21093750e-01,  1.03845772e-02,  2.27244020e+01]),\n",
       " array([-5.56979309e+02,  1.25683618e+00,  1.18947077e+00, ...,\n",
       "         1.56250000e-01,  1.38419535e-02,  2.31243297e+01]),\n",
       " array([-2.24130554e+02,  2.58992462e+01, -1.65767536e+01, ...,\n",
       "         9.57031250e-02,  2.13000009e-02,  2.19643295e+01]),\n",
       " array([-6.94575119e+01,  1.02258316e+02, -2.84007988e+01, ...,\n",
       "         4.19921875e-02,  5.80101381e-03,  1.80183947e+01]),\n",
       " array([-1.59569901e+02,  1.63110840e+02,  2.65717201e+01, ...,\n",
       "         4.39453125e-02,  2.29741285e-02,  9.86797735e+00]),\n",
       " array([-3.74057495e+02,  5.01437721e+01, -4.78473511e+01, ...,\n",
       "         1.58203125e-01,  8.86653819e-03,  1.60717294e+01]),\n",
       " array([-5.64682617e+02,  1.70138226e+01,  2.20570588e+00, ...,\n",
       "         3.22265625e-02,  9.86391098e-03,  3.15048665e+01]),\n",
       " array([-4.92465515e+02,  7.01206512e+01, -5.97079163e+01, ...,\n",
       "         2.49023438e-02,  1.03316761e-02,  1.59626111e+01]),\n",
       " array([-4.99629364e+02,  7.79119110e+01,  4.84273720e+00, ...,\n",
       "         7.22656250e-02,  1.02626542e-02,  1.80575952e+01]),\n",
       " array([-1.43240128e+02,  1.19252609e+02,  2.58339424e+01, ...,\n",
       "         1.42089844e-01,  2.83933982e-02,  2.21303672e+01]),\n",
       " array([-6.59909180e+02,  1.21479816e+01,  7.63055801e+00, ...,\n",
       "         1.79199219e-01,  1.44748764e-02,  2.09022492e+01]),\n",
       " array([-1.76826279e+02,  1.02174820e+02,  1.28314781e+01, ...,\n",
       "         4.49218750e-02,  8.78990932e-03,  2.27999717e+01]),\n",
       " array([-5.48711121e+02,  7.37690506e+01,  1.34664040e+01, ...,\n",
       "         5.95703125e-02,  7.37220237e-03,  1.41296059e+01]),\n",
       " array([-5.50761414e+02,  1.56142483e+01,  1.38484325e+01, ...,\n",
       "         3.27148438e-02,  4.64604844e-03,  2.01899797e+01]),\n",
       " array([-3.06185394e+02,  1.24938965e+02,  1.05137386e+01, ...,\n",
       "         5.66406250e-02,  1.98526587e-02,  1.58342585e+01]),\n",
       " array([-6.25567566e+02,  1.66211205e+01,  1.12768154e+01, ...,\n",
       "         1.32812500e-01,  1.55039157e-02,  1.46673903e+01]),\n",
       " array([-5.88060852e+02,  4.34318771e+01,  8.89430046e+00, ...,\n",
       "         2.20214844e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-9.06557236e+01,  9.59063034e+01,  2.40182114e+01, ...,\n",
       "         1.39648438e-01,  1.24385801e-02,  2.16517687e+01]),\n",
       " array([-1.94189453e+02,  9.90595016e+01, -3.66999588e+01, ...,\n",
       "         1.66015625e-02,  9.11273665e-03,  2.45373547e+01]),\n",
       " array([-4.09503113e+02,  6.45826416e+01,  1.69678154e+01, ...,\n",
       "         7.86132812e-02,  4.23275407e-02,  1.59871474e+01]),\n",
       " array([-2.14400223e+02,  1.13964859e+02, -1.14673767e+01, ...,\n",
       "         3.07617188e-02,  1.94155839e-02,  1.45571709e+01]),\n",
       " array([-2.65379883e+02,  1.31255005e+02, -3.81277084e+01, ...,\n",
       "         1.42089844e-01,  8.87871304e-03,  1.58774972e+01]),\n",
       " array([-5.97493408e+02,  6.33058624e+01,  2.77803802e+01, ...,\n",
       "         1.41113281e-01,  1.44818734e-02,  2.08053991e+01]),\n",
       " array([-3.36274292e+02,  1.29484665e+02,  3.77822876e+01, ...,\n",
       "         2.92968750e-02,  2.06266634e-02,  1.41260982e+01]),\n",
       " array([-4.66220734e+02,  3.20034332e+01, -4.37788162e+01, ...,\n",
       "         1.51367188e-01,  7.18810201e-03,  1.54405341e+01]),\n",
       " array([-9.75742493e+01,  7.93807831e+01, -5.09189873e+01, ...,\n",
       "         1.55273438e-01,  8.98054771e-03,  2.10972622e+01]),\n",
       " array([-5.20649292e+02,  6.33330460e+01,  3.15687981e+01, ...,\n",
       "         3.07617188e-02,  3.48349395e-02,  2.48224020e+01]),\n",
       " array([-1.96984116e+02, -4.71144753e+01, -2.29146385e+01, ...,\n",
       "         1.85546875e-02,  2.30155779e-02,  1.28163368e+01]),\n",
       " array([-2.73892517e+02,  1.19375198e+02,  1.76194077e+01, ...,\n",
       "         1.01074219e-01,  2.14832617e-02,  1.88302338e+01]),\n",
       " array([-1.82229187e+02,  1.36773346e+02, -1.97324123e+01, ...,\n",
       "         1.66015625e-02,  6.67109511e-03,  1.68406293e+01]),\n",
       " array([-4.21223022e+02,  9.06214905e+01,  3.88566685e+00, ...,\n",
       "         4.68750000e-02,  2.21281445e-02,  1.46490023e+01]),\n",
       " array([-3.25020691e+02,  8.68274307e+01,  1.60860329e+01, ...,\n",
       "         2.34375000e-02,  1.64443083e-02,  1.05702110e+01]),\n",
       " array([-5.22322144e+02,  7.85465698e+01,  2.59422035e+01, ...,\n",
       "         3.02734375e-02,  1.34567362e-02,  1.72253475e+01]),\n",
       " array([-2.31681213e+02,  1.04330795e+02,  1.94989014e+01, ...,\n",
       "         2.53906250e-02,  1.66943985e-02,  1.29632652e+01]),\n",
       " array([-6.06825256e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         2.31933594e-01,  1.64704668e-02,  1.15204155e+01]),\n",
       " array([-5.10137482e+02,  9.36120415e+00,  6.77603817e+00, ...,\n",
       "         2.94921875e-01,  1.29491653e-02,  1.41447142e+01]),\n",
       " array([-1.44200653e+02,  8.25772476e+01, -5.50160408e+01, ...,\n",
       "         2.76855469e-01,  5.77931481e-03,  2.66389370e+01]),\n",
       " array([-5.79710266e+02,  1.59542084e+01,  1.26650562e+01, ...,\n",
       "         1.02539062e-01,  1.70280621e-02,  1.63240701e+01]),\n",
       " array([-3.60809967e+02,  3.51226730e+01, -1.53837311e+02, ...,\n",
       "         2.19726562e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-3.84515198e+02,  1.31092499e+02, -2.63752899e+01, ...,\n",
       "         1.07421875e-01,  6.16510632e-03,  2.08923101e+01]),\n",
       " array([-4.43062897e+02,  1.15228485e+02,  1.64994984e+01, ...,\n",
       "         3.02734375e-02,  3.36878945e-02,  1.89310759e+01]),\n",
       " array([-3.62076782e+02, -1.15018799e+02,  1.48729553e+01, ...,\n",
       "         7.42187500e-02,  1.49280205e-02,  1.61666721e+01]),\n",
       " array([-4.65217072e+02,  3.12162838e+01, -9.21351547e+01, ...,\n",
       "         1.89941406e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.89060699e+02,  9.94447937e+01,  1.48648148e+01, ...,\n",
       "         1.61621094e-01,  2.61950159e-02,  2.41644800e+01]),\n",
       " array([-2.90086792e+02,  9.55545502e+01, -3.12401791e+01, ...,\n",
       "         6.00585938e-02,  6.75945067e-03,  1.59773111e+01]),\n",
       " array([-3.94531555e+02,  1.05038589e+02,  1.63689632e+01, ...,\n",
       "         8.78906250e-03,  3.34707733e-02,  1.89394176e+01]),\n",
       " array([-1.08488373e+02,  8.88866653e+01,  1.91239967e+01, ...,\n",
       "         7.03125000e-02,  2.81703472e-02,  1.31349415e+01]),\n",
       " array([-4.33456116e+02,  4.82033653e+01,  4.25987530e+00, ...,\n",
       "         7.61718750e-02,  7.72606206e-03,  1.77578643e+01]),\n",
       " array([-1.33710297e+02,  1.03842697e+02, -4.25377884e+01, ...,\n",
       "         3.66210938e-02,  1.15605949e-02,  2.14372560e+01]),\n",
       " array([-2.34609573e+02,  8.86659851e+01, -3.09609165e+01, ...,\n",
       "         3.71093750e-02,  1.68882124e-02,  3.39804113e+01]),\n",
       " array([-5.38147095e+02,  2.08565331e+01,  1.73517895e+01, ...,\n",
       "         1.47460938e-01,  2.59248483e-02,  1.40109301e+01]),\n",
       " array([-5.06019745e+02,  6.56075287e+01, -3.08362198e+00, ...,\n",
       "         4.63867188e-02,  1.02321852e-02,  1.96399420e+01]),\n",
       " array([-4.51485657e+02,  3.43809128e+01,  2.17394600e+01, ...,\n",
       "         7.61718750e-02,  8.36819888e-03,  1.62443653e+01]),\n",
       " array([-2.42995514e+02,  1.47638214e+02,  1.38057232e+01, ...,\n",
       "         6.34765625e-02,  4.40061384e-02,  1.77132055e+01]),\n",
       " array([-6.00642212e+02,  9.70596790e+00,  4.56380129e+00, ...,\n",
       "         1.08886719e-01,  1.79164555e-02,  1.60668686e+01]),\n",
       " array([-6.02297546e+02,  2.71860046e+01,  2.10502396e+01, ...,\n",
       "         1.46484375e-01,  4.90026483e-02,  1.40585870e+01]),\n",
       " array([-2.16169022e+02,  1.17614304e+02,  2.41289577e+01, ...,\n",
       "         5.81054688e-02,  1.15342593e-02,  1.56253010e+01]),\n",
       " array([-6.41102173e+02,  5.02277451e+01,  3.19703617e+01, ...,\n",
       "         5.32226562e-02,  2.54055177e-02,  3.02981585e+01]),\n",
       " array([-6.37066956e+02,  4.78920593e+01,  9.33785200e-01, ...,\n",
       "         1.76757812e-01,  1.32852568e-02,  1.90367475e+01]),\n",
       " array([-2.95214600e+02,  1.85959778e+02,  7.17758255e+01, ...,\n",
       "         3.61328125e-02,  1.77364395e-02,  1.61803499e+01]),\n",
       " array([-2.45599579e+02,  1.18869766e+02,  3.13710511e-01, ...,\n",
       "         4.63867188e-02,  1.46300532e-02,  1.30843654e+01]),\n",
       " array([-1.71414200e+02,  9.32068481e+01, -2.75532131e+01, ...,\n",
       "         7.22656250e-02,  1.80343642e-02,  2.09212258e+01]),\n",
       " array([-3.99026764e+02,  1.03538483e+02, -4.62030029e+00, ...,\n",
       "         9.03320312e-02,  7.91515162e-03,  1.38331950e+01]),\n",
       " array([-1.78921616e+02,  1.44118668e+02, -5.56368675e+01, ...,\n",
       "         9.66796875e-02,  2.35354024e-02,  1.42760083e+01]),\n",
       " array([-4.46239441e+02,  5.08767853e+01,  6.11955032e+01, ...,\n",
       "         1.70898438e-02,  1.99176553e-02,  2.06539571e+01]),\n",
       " array([-5.37456909e+02,  3.59555130e+01,  4.07833786e+01, ...,\n",
       "         3.51562500e-02,  5.99031085e-03,  1.65151492e+01]),\n",
       " array([-2.63360809e+02,  1.35201416e+02, -1.12364101e+01, ...,\n",
       "         4.49218750e-02,  1.23345232e-02,  1.08646803e+01]),\n",
       " array([-5.35117615e+02,  3.16848621e+01,  1.24839325e+01, ...,\n",
       "         1.78222656e-01,  1.36166380e-02,  2.31874093e+01]),\n",
       " array([-4.24503601e+02,  1.04563660e+02,  2.55416641e+01, ...,\n",
       "         4.19921875e-02,  9.25063657e-03,  1.57085761e+01]),\n",
       " array([-1.45498688e+02,  7.21616745e+01, -1.86981926e+01, ...,\n",
       "         1.07910156e-01,  4.04182442e-03,  2.27394894e+01]),\n",
       " array([-6.26799988e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         3.72558594e-01,  2.06389976e-02,  1.70329615e+01]),\n",
       " array([-5.57819458e+02,  5.32243385e+01,  3.21901588e+01, ...,\n",
       "         1.05957031e-01,  5.98977565e-04,  1.28668404e+02]),\n",
       " array([-2.67952209e+02,  8.74094543e+01,  1.06107826e+01, ...,\n",
       "         2.97851562e-02,  3.93172051e-02,  2.32850939e+01]),\n",
       " array([-7.16262146e+02,  4.57027316e-01,  4.36341703e-01, ...,\n",
       "         1.36718750e-02,  4.61564067e-03,  6.02765679e+00]),\n",
       " array([-1.15793503e+02,  7.85229568e+01, -2.60389137e+01, ...,\n",
       "         1.43554688e-01,  1.30728558e-02,  1.79091692e+01]),\n",
       " array([-2.44117386e+02,  4.40495262e+01, -5.87942276e+01, ...,\n",
       "         2.44140625e-02,  1.10546013e-02,  5.47013819e+01]),\n",
       " array([-4.37834229e+02,  9.33922882e+01,  1.26921234e+01, ...,\n",
       "         4.00390625e-02,  2.53546823e-02,  6.54635429e+00]),\n",
       " array([-6.03601990e+02,  3.41051178e+01,  1.54303379e+01, ...,\n",
       "         2.78320312e-02,  2.14130717e-02,  1.53973401e+01]),\n",
       " array([-4.48606750e+02,  9.65280609e+01,  5.38602829e+00, ...,\n",
       "         8.34960938e-02,  2.48135010e-02,  1.74456507e+01]),\n",
       " array([-5.59487305e+02,  2.61344566e+01,  1.52811165e+01, ...,\n",
       "         4.54101562e-02,  4.51707631e-03,  1.53850064e+01]),\n",
       " array([-5.39990967e+02,  4.32039146e+01, -9.62451363e+00, ...,\n",
       "         7.22656250e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.88238464e+02,  1.09074005e+02,  5.44048738e+00, ...,\n",
       "         7.12890625e-02,  1.20768582e-02,  1.96566924e+01]),\n",
       " array([-1.97433853e+01,  1.50950562e+02,  3.60655260e+00, ...,\n",
       "         1.66015625e-02,  1.29035021e-02,  8.97662118e+00]),\n",
       " array([-3.85109940e+01,  8.33923950e+01,  1.36622467e+01, ...,\n",
       "         5.46875000e-02,  1.77942906e-02,  2.53585726e+01]),\n",
       " array([-4.64118835e+02,  5.32486725e+01, -2.70886917e+01, ...,\n",
       "         4.56542969e-01,  2.06268443e-02,  1.68639287e+01]),\n",
       " array([-2.57868439e+02, -1.30587097e+02,  1.74815407e+01, ...,\n",
       "         5.51757812e-02,  2.72539290e-02,  1.44448370e+01]),\n",
       " array([-4.07436615e+02,  1.20908142e+02, -1.07981472e+01, ...,\n",
       "         1.55761719e-01,  1.89758066e-02,  1.91242188e+01]),\n",
       " array([-3.66595123e+02,  8.40100937e+01,  3.30401459e+01, ...,\n",
       "         5.26855469e-01,  9.52660012e-03,  1.03066795e+01]),\n",
       " array([-1.75953156e+02,  1.34733246e+02, -3.82794762e+01, ...,\n",
       "         5.71289062e-02,  1.48281530e-02,  1.63635641e+01]),\n",
       " array([-4.09983978e+02,  3.82760286e+00,  9.67409325e+00, ...,\n",
       "         1.70898438e-01,  7.65048536e-03,  1.76421404e+01]),\n",
       " array([-5.00596985e+02, -1.32471144e+00, -2.68099289e+01, ...,\n",
       "         1.52832031e-01,  1.03841914e-02,  2.34357774e+01]),\n",
       " array([-3.25192413e+02,  5.54103546e+01, -2.65366554e+01, ...,\n",
       "         2.07031250e-01,  1.64557932e-02,  1.09927982e+01]),\n",
       " array([-2.34040085e+02,  4.30013046e+01,  2.40957413e+01, ...,\n",
       "         2.97851562e-02,  2.14319444e-02,  1.54889703e+01]),\n",
       " array([-2.97480682e+02,  1.08949692e+02,  3.50798073e+01, ...,\n",
       "         6.64062500e-02,  2.39567447e-02,  1.84328958e+01]),\n",
       " array([-5.45220154e+02,  1.77892342e+01, -1.93061867e+01, ...,\n",
       "         1.36718750e-01,  1.20434310e-02,  2.12818921e+01]),\n",
       " array([-2.37869614e+02,  1.41792618e+02, -7.91949558e+00, ...,\n",
       "         2.68066406e-01,  4.17658748e-03,  1.28098145e+01]),\n",
       " array([-5.07455872e+02,  4.06902771e+01, -5.00168762e+01, ...,\n",
       "         1.04003906e-01,  2.11764210e-02,  2.68232018e+01]),\n",
       " array([-4.45128815e+02,  1.24347290e+02,  3.42729797e+01, ...,\n",
       "         2.68554688e-02,  1.94440002e-02,  1.42874226e+01]),\n",
       " array([-3.84890503e+02,  1.47233032e+02,  3.68225403e+01, ...,\n",
       "         7.08007812e-02,  1.37734550e-02,  1.40392810e+01]),\n",
       " array([-1.62536438e+02,  1.00632256e+02, -6.20439072e+01, ...,\n",
       "         5.66406250e-02,  8.33702706e-03,  1.51387468e+01]),\n",
       " array([-2.54696609e+02,  2.02543060e+02, -5.23188705e+01, ...,\n",
       "         6.00585938e-02,  2.21041356e-02,  2.05570668e+01]),\n",
       " array([-3.39551514e+02,  1.30452271e+02, -2.22916126e+01, ...,\n",
       "         5.07812500e-02,  2.67234644e-02,  2.81081051e+01]),\n",
       " array([-1.94874054e+02,  1.04275970e+02, -1.62249413e+01, ...,\n",
       "         1.97265625e-01,  1.07013972e-02,  1.92922413e+01]),\n",
       " array([-1.87378769e+02,  1.30153442e+02, -1.36947594e+01, ...,\n",
       "         1.38183594e-01,  3.19310776e-02,  1.52231753e+01]),\n",
       " array([-6.53098083e+02,  4.14150162e+01, -1.65629177e+01, ...,\n",
       "         1.39648438e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.06058441e+02,  1.75972855e+02, -3.13946295e+00, ...,\n",
       "         1.08398438e-01,  2.12077505e-02,  1.51473224e+01]),\n",
       " array([-1.62474701e+02,  9.87049103e+01,  1.41262567e+00, ...,\n",
       "         4.44335938e-01,  1.91692542e-02,  1.67879924e+01]),\n",
       " array([-1.86195190e+02,  1.30470123e+02, -2.27284698e+01, ...,\n",
       "         4.83398438e-02,  1.70792525e-02,  1.87876552e+01]),\n",
       " array([-8.22947083e+01,  8.20277786e+01,  2.87098713e+01, ...,\n",
       "         1.45996094e-01,  2.13237062e-02,  1.64128557e+01]),\n",
       " array([-5.60147522e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         6.10351562e-02,  5.55493718e-03,  7.55863935e+00]),\n",
       " array([-5.81073547e+02,  9.08609581e+00,  5.89503765e+00, ...,\n",
       "         9.91210938e-02,  8.57457095e-03,  9.69733372e+00]),\n",
       " array([-1.76570801e+02,  7.50915604e+01, -8.43237000e+01, ...,\n",
       "         1.47949219e-01,  1.44452131e-02,  2.10168242e+01]),\n",
       " array([-3.66491486e+02,  1.53000164e+01,  8.00820465e+01, ...,\n",
       "         4.19921875e-02,  1.26501041e-02,  1.89565554e+01]),\n",
       " array([-1.58564621e+02,  9.70500183e+01, -7.38078003e+01, ...,\n",
       "         6.59179688e-02,  6.99424657e-03,  1.19368605e+01]),\n",
       " array([-3.79446503e+02,  1.77677795e+02,  4.86589813e+00, ...,\n",
       "         5.85937500e-03,  1.28566484e-02,  1.50020689e+01]),\n",
       " array([-3.35354156e+02, -2.00128212e+01, -3.02895451e+01, ...,\n",
       "         4.24804688e-02,  2.73727029e-02,  2.50555009e+01]),\n",
       " array([-4.76091705e+02,  1.86797023e+00, -1.99374008e+00, ...,\n",
       "         4.29687500e-02,  8.17483514e-03,  1.99442163e+01]),\n",
       " array([-2.35911819e+02,  1.21289658e+02, -3.93757858e+01, ...,\n",
       "         4.73632812e-02,  1.25882546e-02,  1.89693332e+01]),\n",
       " array([-5.26820190e+02,  5.68646393e+01,  3.59556503e+01, ...,\n",
       "         1.46484375e-02,  8.43142619e-03,  1.17270835e+01]),\n",
       " array([-2.42833298e+02,  1.51852173e+02, -9.96556091e+01, ...,\n",
       "         5.07812500e-02,  1.66183322e-02,  1.08448610e+01]),\n",
       " array([-1.29360580e+02, -8.22927322e+01,  2.04408283e+01, ...,\n",
       "         1.10351562e-01,  2.40188386e-02,  2.11362630e+01]),\n",
       " array([-2.71622742e+02,  4.04816437e+01,  2.56124077e+01, ...,\n",
       "         1.03515625e-01,  2.78453958e-03,  7.91197941e+00]),\n",
       " array([-3.73118469e+02,  3.22570114e+01, -1.15884201e+02, ...,\n",
       "         2.70507812e-01,  1.19693303e-02,  1.96341053e+01]),\n",
       " array([-2.73259705e+02, -9.12887573e+00,  6.99898148e+01, ...,\n",
       "         4.29687500e-02,  7.92565142e-03,  1.10596895e+01]),\n",
       " array([-5.01463989e+02, -3.33712816e-01,  4.12928772e+01, ...,\n",
       "         5.07812500e-02,  9.97665213e-03,  1.49529278e+01]),\n",
       " array([-5.44601624e+02,  2.54927444e+01,  1.14976149e+01, ...,\n",
       "         5.32226562e-02,  2.13641215e-02,  1.42687812e+01]),\n",
       " array([-1.64149567e+02,  1.20130371e+02,  6.32426643e+00, ...,\n",
       "         8.10546875e-02,  1.09452735e-02,  1.55610204e+01]),\n",
       " array([-6.45130066e+02,  3.55657463e+01,  2.80207539e+00, ...,\n",
       "         5.17578125e-02,  1.32696815e-02,  2.11037830e+01]),\n",
       " array([-9.09580612e+01,  6.14171600e+01, -6.71648712e+01, ...,\n",
       "         4.34570312e-02,  1.53873061e-02,  8.52799639e+00]),\n",
       " array([-1.25967422e+02,  1.18749588e+02, -4.68845520e+01, ...,\n",
       "         6.25000000e-02,  1.19791168e-02,  3.02539110e+01]),\n",
       " array([-1.81252640e+02,  1.01451157e+02, -5.17194290e+01, ...,\n",
       "         7.22656250e-02,  1.48375750e-02,  2.28313118e+01]),\n",
       " array([-2.05868912e+02,  1.76620422e+02,  1.24043202e+00, ...,\n",
       "         1.75781250e-02,  2.14525366e-02,  1.07035421e+01]),\n",
       " array([-7.73849487e+01,  1.00205887e+02, -5.34748917e+01, ...,\n",
       "         2.19726562e-01,  1.09468525e-02,  1.52474791e+01]),\n",
       " array([-1.04813560e+02,  1.15373177e+02, -6.97344818e+01, ...,\n",
       "         1.10839844e-01,  1.92153212e-02,  2.10314527e+01]),\n",
       " array([-2.64735687e+02,  7.03294373e+01,  3.89539337e+01, ...,\n",
       "         7.03125000e-02,  2.64351442e-02,  1.64816692e+01]),\n",
       " array([-4.00826874e+02,  1.49951141e+02,  2.11129837e+01, ...,\n",
       "         4.88281250e-02,  5.34657828e-03,  1.01572484e+01]),\n",
       " array([-5.66804565e+02,  1.28955650e+01, -8.60180569e+00, ...,\n",
       "         1.61132812e-01,  1.12072569e-02,  2.30410740e+01]),\n",
       " array([-6.36501343e+02,  5.43035202e+01,  3.09994526e+01, ...,\n",
       "         5.32226562e-02,  2.22159996e-02,  1.87849358e+01]),\n",
       " array([-1.95762833e+02,  1.84971542e+02, -1.55266361e+01, ...,\n",
       "         3.28125000e-01,  1.51198650e-02,  1.20570548e+01]),\n",
       " array([-3.92167175e+02,  1.11344421e+02,  3.04942856e+01, ...,\n",
       "         1.12792969e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-4.51505768e+02,  9.06858521e+01, -1.58826160e+00, ...,\n",
       "         8.59375000e-02,  1.36994158e-02,  1.46002322e+01]),\n",
       " array([-1.46053253e+02,  1.17636398e+02, -7.57690735e+01, ...,\n",
       "         1.51855469e-01,  1.18049857e-02,  2.49009088e+01]),\n",
       " array([-5.10997681e+02,  4.65416183e+01,  2.80544510e+01, ...,\n",
       "         2.34375000e-02,  3.44177769e-02,  2.67651409e+01]),\n",
       " array([-2.84983978e+02,  1.28545593e+02,  1.97151108e+01, ...,\n",
       "         1.09375000e-01,  1.30595149e-02,  2.38663420e+01]),\n",
       " array([-5.01142853e+02,  7.31482792e+00,  5.11161613e+00, ...,\n",
       "         5.85937500e-02,  1.04807977e-02,  2.24231899e+01]),\n",
       " array([-5.20203369e+02,  3.24780121e+01,  2.88187561e+01, ...,\n",
       "         8.54492188e-02,  1.21466119e-02,  2.30596289e+01]),\n",
       " array([-9.42690506e+01,  3.39933357e+01,  3.56646276e+00, ...,\n",
       "         2.66601562e-01,  1.07276248e-02,  1.79692000e+01]),\n",
       " array([-5.27474487e+02,  6.23122711e+01,  4.58279037e+01, ...,\n",
       "         2.68554688e-02,  4.67424560e-02,  2.26392895e+01]),\n",
       " array([-2.27662003e+02, -4.45138092e+01,  4.05782700e+01, ...,\n",
       "         5.12695312e-02,  3.77564990e-02,  1.60928354e+01]),\n",
       " array([-6.25700912e+01,  1.18522705e+02, -3.57440567e+01, ...,\n",
       "         2.73437500e-02,  2.97778716e-02,  1.37562990e+01]),\n",
       " array([-2.17559494e+02,  4.20451126e+01, -7.07789612e+01, ...,\n",
       "         1.85058594e-01,  1.03501759e-02,  2.32796907e+01]),\n",
       " array([-1.68173355e+02,  1.51725586e+02, -5.52110329e+01, ...,\n",
       "         1.11328125e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.74002304e+02,  1.61390244e+02, -4.08268204e+01, ...,\n",
       "         1.05957031e-01,  1.65813852e-02,  2.38712147e+01]),\n",
       " array([-5.15631409e+02,  5.21822128e+01,  1.72257233e+01, ...,\n",
       "         8.30078125e-02,  1.41775123e-02,  1.71810314e+01]),\n",
       " array([-2.65567627e+02,  1.36795197e+02,  1.73570404e+01, ...,\n",
       "         7.22656250e-02,  2.23180596e-02,  1.90089837e+01]),\n",
       " array([-2.53606140e+02,  1.19417946e+02,  2.69583626e+01, ...,\n",
       "         8.59375000e-02,  6.07375534e-03,  2.42223307e+01]),\n",
       " array([-9.69459839e+01,  8.70938797e+01, -2.81458130e+01, ...,\n",
       "         8.00781250e-02,  2.42459534e-02,  1.54764846e+01]),\n",
       " array([-3.94099304e+02,  1.11878723e+02,  2.06035881e+01, ...,\n",
       "         7.76367188e-02,  1.39290479e-02,  1.86164021e+01]),\n",
       " array([-1.97274414e+02,  1.34038467e+02,  6.29481125e+00, ...,\n",
       "         1.85546875e-02,  2.10140521e-02,  1.03351049e+01]),\n",
       " array([-1.25626595e+02,  1.40829651e+02, -9.05950623e+01, ...,\n",
       "         8.15429688e-02,  1.07746565e-02,  1.45134240e+01]),\n",
       " array([-4.70643829e+02, -1.69090424e+01, -5.64728355e+01, ...,\n",
       "         9.08203125e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.69540848e+02,  1.48008087e+02,  1.26262760e+01, ...,\n",
       "         1.51367188e-02,  1.56852576e-02,  1.47982255e+01]),\n",
       " array([-2.29059647e+02,  1.10811279e+02, -3.11966400e+01, ...,\n",
       "         1.13769531e-01,  1.33205865e-02,  2.17803180e+01]),\n",
       " array([-1.71738373e+02,  5.16555328e+01, -2.43885002e+01, ...,\n",
       "         1.72363281e-01,  1.15626403e-02,  3.45717758e+01]),\n",
       " array([-7.76367645e+01,  9.31152878e+01, -1.40672379e+01, ...,\n",
       "         1.46484375e-02,  1.61627802e-02,  1.36133268e+01]),\n",
       " array([-3.22943512e+02,  1.17290611e+02, -2.81392479e+00, ...,\n",
       "         1.57714844e-01,  7.41939239e-03,  1.98185146e+01]),\n",
       " array([-6.30151001e+02,  2.47899952e+01,  3.22893572e+00, ...,\n",
       "         5.51757812e-02,  5.35187227e-03,  9.13248360e+00]),\n",
       " array([-3.63665314e+02,  1.16503677e+01,  4.25291595e+01, ...,\n",
       "         6.15234375e-02,  5.25767226e-02,  1.36019692e+01]),\n",
       " array([-5.23012573e+02,  6.80149384e+01,  1.02549706e+01, ...,\n",
       "         9.32617188e-02,  2.17488266e-02,  9.60265175e+00]),\n",
       " array([-2.81378632e+02,  1.17894745e+02, -3.64698296e+01, ...,\n",
       "         1.07910156e-01,  1.62561074e-02,  3.25066358e+01]),\n",
       " array([-4.62521881e+02,  2.49687786e+01, -3.12529850e+00, ...,\n",
       "         1.11328125e-01,  5.93022628e-03,  1.46416232e+01]),\n",
       " array([-4.32619537e+02,  1.12251694e+02, -2.61671028e+01, ...,\n",
       "         1.91894531e-01,  1.65581674e-02,  1.99659795e+01]),\n",
       " array([-1.71310089e+02,  9.77325897e+01,  1.34916363e+01, ...,\n",
       "         6.88476562e-02,  1.34772339e-02,  1.32077113e+01]),\n",
       " array([-5.01356812e+02,  9.19358826e+01,  5.55896873e+01, ...,\n",
       "         5.56640625e-02,  2.12472921e-02,  1.97671339e+01]),\n",
       " array([-6.10617920e+02,  5.67414665e+01,  1.77346077e+01, ...,\n",
       "         7.81250000e-02,  1.23832120e-02,  1.70644030e+01]),\n",
       " array([-3.48098511e+02, -7.91502686e+01, -5.11535721e+01, ...,\n",
       "         1.18164062e-01,  6.09885091e-03,  2.32112929e+01]),\n",
       " array([-2.72990173e+02, -5.88317776e+00,  9.80334663e+00, ...,\n",
       "         1.98730469e-01,  3.72477666e-02,  2.55215645e+01]),\n",
       " array([-3.90721130e+02,  5.98370514e+01, -6.86395111e+01, ...,\n",
       "         4.15039062e-02,  2.09651905e-02,  1.35214061e+01]),\n",
       " array([-3.63708382e+01,  1.23525124e+01, -3.37583008e+01, ...,\n",
       "         1.66503906e-01,  5.22737238e-03,  2.21201569e+01]),\n",
       " array([-1.15161385e+02,  7.43967285e+01, -2.08890533e+01, ...,\n",
       "         1.84082031e-01,  2.13573521e-02,  1.22736819e+01]),\n",
       " array([-2.07015091e+02,  1.07944046e+02, -1.27412844e+01, ...,\n",
       "         9.08203125e-02,  1.95283031e-02,  2.34697834e+01]),\n",
       " array([-4.75595306e+02,  8.11677094e+01, -2.36224632e+01, ...,\n",
       "         4.10156250e-02,  3.28686786e-02,  1.95462495e+01]),\n",
       " array([-4.94056244e+02,  8.98512039e+01, -1.39653721e+01, ...,\n",
       "         6.05468750e-02,  3.88944140e-02,  2.61687875e+01]),\n",
       " array([-4.35426727e+02,  1.43868317e+02,  2.50173626e+01, ...,\n",
       "         1.94335938e-01,  7.94570710e-03,  1.27336860e+01]),\n",
       " array([-2.88912933e+02, -6.43129044e+01,  1.46285229e+01, ...,\n",
       "         4.54101562e-02,  1.16157444e-02,  2.38672674e+01]),\n",
       " array([-2.01472946e+02,  2.15310272e+02, -6.86291046e+01, ...,\n",
       "         5.85937500e-02,  1.65437299e-02,  1.71159521e+01]),\n",
       " array([-7.36061096e+01,  6.96262512e+01, -7.00057449e+01, ...,\n",
       "         1.51367188e-02,  4.43774226e-02,  2.72176325e+01]),\n",
       " array([-1.52435577e+02,  9.70371399e+01,  1.06638870e+01, ...,\n",
       "         6.29882812e-02,  1.08196490e-02,  2.44776100e+01]),\n",
       " array([-1.64963562e+02,  1.56531952e+02, -4.35325928e+01, ...,\n",
       "         7.66601562e-02,  4.02926261e-02,  1.66753516e+01]),\n",
       " array([-6.04499695e+02, -1.66165903e-02, -2.23538771e-01, ...,\n",
       "         2.83203125e-02,  2.37084696e-02,  1.17623068e+01]),\n",
       " array([-3.98014191e+02,  4.79491463e+01, -2.42114143e+01, ...,\n",
       "         1.05468750e-01,  1.92370643e-02,  4.44848686e+01]),\n",
       " array([-2.40703384e+02,  7.24247131e+01, -4.79161758e+01, ...,\n",
       "         1.57714844e-01,  1.28631961e-02,  2.29641914e+01]),\n",
       " array([-2.45478836e+02,  1.36977249e+02,  2.33427067e+01, ...,\n",
       "         1.11816406e-01,  1.94251910e-02,  1.81335256e+01]),\n",
       " array([-1.76026337e+02,  1.23788284e+02,  3.82299042e+01, ...,\n",
       "         3.61328125e-02,  2.04307827e-02,  1.09083116e+01]),\n",
       " array([-1.67403687e+02,  1.66235153e+02, -6.55989685e+01, ...,\n",
       "         1.28417969e-01,  3.58809698e-02,  1.37246072e+01]),\n",
       " array([-4.33911591e+02, -1.54958458e+01, -2.37624139e-01, ...,\n",
       "         1.30371094e-01,  8.30826938e-03,  1.55073017e+01]),\n",
       " array([-6.18724899e+01,  1.40412018e+02, -6.00831223e+01, ...,\n",
       "         6.59179688e-02,  1.75626203e-02,  1.20549716e+01]),\n",
       " array([-2.18634979e+02,  1.40623062e+02,  2.72517929e+01, ...,\n",
       "         6.78710938e-02,  1.62927883e-02,  1.59719482e+01]),\n",
       " array([-5.00445709e+02,  3.50552225e+00, -9.51363182e+00, ...,\n",
       "         7.66601562e-02,  1.20945471e-02,  2.62999684e+01]),\n",
       " array([-7.02202454e+02,  4.73803558e+01, -3.66745453e+01, ...,\n",
       "         6.73828125e-02,  1.51476156e-02,  1.93180963e+01]),\n",
       " array([-5.73521362e+02,  3.81646690e+01,  2.83353472e+00, ...,\n",
       "         7.32421875e-02,  1.90261831e-02,  1.71933994e+01]),\n",
       " array([-5.16720276e+02,  5.85905075e+01,  2.10453224e+01, ...,\n",
       "         7.37304688e-02,  1.03430908e-02,  1.53572395e+01]),\n",
       " array([-8.25269409e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         1.89941406e-01,  1.79676043e-02,  1.88701212e+01]),\n",
       " array([-1.55203842e+02,  1.46518402e+02, -1.77660255e+01, ...,\n",
       "         1.78710938e-01,  1.33990715e-02,  9.71951112e+00]),\n",
       " array([-4.29596527e+02, -8.06548786e+00, -9.94757843e+01, ...,\n",
       "         1.66503906e-01,  7.46299489e-03,  1.32804334e+01]),\n",
       " array([-3.02942078e+02,  1.22252495e+02, -3.12443104e+01, ...,\n",
       "         2.15332031e-01,  1.36392172e-02,  1.88011304e+01]),\n",
       " array([-1.80198242e+02,  2.74277496e+01, -6.81256256e+01, ...,\n",
       "         4.02343750e-01,  1.23511489e-02,  1.80459037e+01]),\n",
       " array([-2.31758469e+02,  1.28998459e+02, -1.26034832e+00, ...,\n",
       "         1.31835938e-02,  2.78159427e-02,  2.30057523e+01]),\n",
       " array([-2.41284866e+02,  1.32224091e+02, -5.91774483e+01, ...,\n",
       "         1.34277344e-01,  1.31991070e-02,  1.49737924e+01]),\n",
       " array([-5.87640869e+02,  6.95204010e+01,  3.52491713e+00, ...,\n",
       "         8.78906250e-02,  7.18598857e-03,  1.02373272e+01]),\n",
       " array([-2.50119904e+02,  1.10220268e+02, -1.21265421e+01, ...,\n",
       "         1.91406250e-01,  1.26868567e-02,  2.41467595e+01]),\n",
       " array([-1.74783524e+02,  1.25606277e+02, -8.85871582e+01, ...,\n",
       "         5.85937500e-02,  7.02201423e-03,  1.72403887e+01]),\n",
       " array([-2.95052979e+02,  1.26294571e+02,  1.70329895e+01, ...,\n",
       "         2.53906250e-02,  1.06163304e-02,  7.61712790e+00]),\n",
       " array([-2.18739532e+02, -2.62633553e+01, -9.24031639e+00, ...,\n",
       "         1.01074219e-01,  8.34043965e-03,  1.64453670e+01]),\n",
       " array([-3.53830078e+02, -7.91910248e+01,  4.57827415e+01, ...,\n",
       "         6.78710938e-02,  1.10659515e-02,  8.20439160e+00]),\n",
       " array([-3.61359650e+02, -5.42962570e+01,  2.95512924e+01, ...,\n",
       "         2.02148438e-01,  1.36145306e-02,  2.19336718e+01]),\n",
       " array([-2.28051102e+02,  1.42688049e+02,  7.28394127e+00, ...,\n",
       "         7.32421875e-02,  1.39155846e-02,  2.29383558e+01]),\n",
       " array([-4.83715546e+02,  7.39319534e+01,  4.82647991e+00, ...,\n",
       "         1.30859375e-01,  4.70825714e-03,  1.54793158e+01]),\n",
       " array([-2.07243240e+02,  1.55747833e+01, -3.06541586e+00, ...,\n",
       "         1.31835938e-01,  1.96454977e-02,  1.67786002e+01]),\n",
       " array([-4.52300690e+02,  6.82104492e+01, -3.94490891e+01, ...,\n",
       "         1.51367188e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-4.73474365e+02,  6.47939301e+01, -2.90732307e+01, ...,\n",
       "         3.95507812e-01,  1.07372938e-02,  2.58061826e+01]),\n",
       " array([-1.84927277e+02, -3.46991234e+01,  7.83547974e+01, ...,\n",
       "         2.97851562e-02,  2.37557604e-02,  1.92040980e+01]),\n",
       " array([-1.30415283e+02,  1.54687103e+02,  4.90721607e+00, ...,\n",
       "         1.31835938e-02,  1.55870037e-02,  1.16504513e+01]),\n",
       " array([-5.51083496e+02,  3.89077568e+00,  3.83218002e+00, ...,\n",
       "         4.83398438e-02,  8.91787016e-03,  1.77595466e+01]),\n",
       " array([-4.86607056e+02,  6.34317207e+01,  2.19640884e+01, ...,\n",
       "         1.52343750e-01,  1.17031836e-02,  1.73717022e+01]),\n",
       " array([-5.20774918e+01,  4.27524261e+01, -3.54881058e+01, ...,\n",
       "         1.40625000e-01,  6.01898122e-03,  2.16909170e+01]),\n",
       " array([-2.78480225e+02,  8.70600586e+01, -5.37648849e+01, ...,\n",
       "         1.83593750e-01,  7.05863105e-03,  2.24789441e+01]),\n",
       " array([-5.21622131e+02, -5.37516327e+01,  4.81704407e+01, ...,\n",
       "         5.58105469e-01,  1.11415533e-02,  2.75833994e+01]),\n",
       " array([-1.19645729e+02,  1.39052399e+02, -2.66091957e+01, ...,\n",
       "         1.54785156e-01,  8.17241211e-03,  1.79581910e+01]),\n",
       " array([-2.16005920e+02,  1.52322357e+02,  2.39751987e+01, ...,\n",
       "         2.63671875e-02,  2.73662908e-02,  1.28972828e+01]),\n",
       " array([-6.02551025e+02,  3.39966049e+01,  2.46576672e+01, ...,\n",
       "         3.58398438e-01,  8.08627361e-03,  2.14370221e+01]),\n",
       " array([-3.52391571e+02,  4.38429642e+01, -1.21725445e+01, ...,\n",
       "         5.61523438e-02,  2.34678418e-02,  1.56092703e+01]),\n",
       " array([-5.13192993e+02,  7.83331451e+01,  3.93096085e+01, ...,\n",
       "         8.88671875e-02,  3.08455973e-02,  3.54308426e+01]),\n",
       " array([-5.14734924e+02,  8.59231567e+01,  5.48037567e+01, ...,\n",
       "         1.51367188e-02,  2.27518230e-02,  1.96389481e+01]),\n",
       " array([-5.47794800e+02, -2.50652933e+00,  2.86035614e+01, ...,\n",
       "         1.32324219e-01,  1.08234106e-02,  1.56926766e+01]),\n",
       " array([-2.10735123e+02,  1.53482742e+02,  2.64645386e+01, ...,\n",
       "         1.15234375e-01,  1.12303107e-02,  1.53746635e+01]),\n",
       " array([-5.22451965e+02,  1.19375219e+01,  6.65244341e+00, ...,\n",
       "         6.54296875e-02,  3.35179070e-02,  2.88603425e+01]),\n",
       " array([-4.70300140e+02,  5.49399834e+01, -2.30195045e+00, ...,\n",
       "         1.03515625e-01,  2.00516490e-02,  1.74605757e+01]),\n",
       " array([-1.42396042e+02,  2.40465889e+01, -9.50782967e+00, ...,\n",
       "         1.12792969e-01,  8.35944432e-03,  2.02001184e+01]),\n",
       " array([-5.68326233e+02,  1.15751934e+00,  1.13524556e+00, ...,\n",
       "         9.76562500e-02,  1.43150939e-02,  1.34386182e+01]),\n",
       " array([-1.94066360e+02,  1.68404892e+02, -5.29893036e+01, ...,\n",
       "         1.93847656e-01,  1.13702700e-02,  1.74422547e+01]),\n",
       " array([-3.99632812e+02,  1.41661652e+02, -4.25584078e+00, ...,\n",
       "         7.61718750e-02,  2.50330342e-02,  1.85524553e+01]),\n",
       " array([-4.61964844e+02,  8.75504684e+01,  3.23278046e+01, ...,\n",
       "         4.44335938e-02,  2.17328531e-02,  1.56877860e+01]),\n",
       " array([-5.73571167e+02,  3.81799316e+01,  1.66118469e+01, ...,\n",
       "         5.81054688e-02,  2.78443494e-02,  1.60503924e+01]),\n",
       " array([-4.48843536e+02,  5.77215652e+01, -3.52431831e+01, ...,\n",
       "         1.00097656e-01,  4.77975305e-03,  1.04548924e+01]),\n",
       " array([-2.89007111e+02,  8.80166321e+01,  1.68271122e+01, ...,\n",
       "         7.42187500e-02,  1.71245274e-02,  2.26562232e+01]),\n",
       " array([-5.44050354e+02,  5.55571861e+01,  7.20121956e+00, ...,\n",
       "         1.25488281e-01,  8.02899876e-03,  2.15456918e+01]),\n",
       " array([-4.75508362e+02,  3.99066772e+01,  3.62462120e+01, ...,\n",
       "         2.58789062e-02,  1.57227080e-02,  2.59923220e+01]),\n",
       " array([-2.01932022e+02,  1.01392838e+02, -6.67155762e+01, ...,\n",
       "         8.49609375e-02,  2.88701685e-02,  1.55929044e+01]),\n",
       " array([-2.08893982e+02,  7.80043640e+01, -2.60400372e+01, ...,\n",
       "         1.13769531e-01,  1.03408853e-02,  2.11204588e+01]),\n",
       " array([-1.26638184e+02,  1.41873108e+02, -4.87466660e+01, ...,\n",
       "         1.46972656e-01,  1.69615325e-02,  1.50950611e+01]),\n",
       " array([-1.55861740e+02,  1.19654190e+02,  3.38501015e+01, ...,\n",
       "         1.85546875e-02,  3.96749728e-02,  2.46413767e+01]),\n",
       " array([-2.25938568e+02,  1.26555222e+02,  9.09631538e+00, ...,\n",
       "         4.63867188e-02,  1.60826481e-02,  1.24824502e+01]),\n",
       " array([-1.40967896e+02,  6.47643661e+01, -9.53349762e+01, ...,\n",
       "         4.19921875e-02,  1.62969904e-02,  1.83078825e+01]),\n",
       " array([-1.63014160e+02,  6.23804321e+01, -5.75143356e+01, ...,\n",
       "         2.34375000e-01,  1.36539551e-02,  1.42368585e+01]),\n",
       " array([-3.46363129e+02,  1.16667236e+02, -1.94072819e+01, ...,\n",
       "         2.65136719e-01,  3.00126156e-02,  2.20025808e+01]),\n",
       " array([-7.54356995e+01,  6.09154358e+01, -7.39579163e+01, ...,\n",
       "         9.42382812e-02,  4.84754204e-03,  6.08486176e+01]),\n",
       " array([-1.88135895e+02,  9.38588867e+01, -2.67523727e+01, ...,\n",
       "         1.90917969e-01,  1.33172332e-02,  1.78387702e+01]),\n",
       " array([-1.76505737e+02,  5.24248886e+01, -1.38141937e+01, ...,\n",
       "         6.93359375e-02,  1.21348645e-02,  1.50865674e+01]),\n",
       " array([-5.12840637e+02,  4.73064270e+01, -2.51358986e+01, ...,\n",
       "         9.47265625e-02,  1.35131764e-02,  2.91858613e+01]),\n",
       " array([-4.63937744e+02,  2.77382584e+01, -4.45154877e+01, ...,\n",
       "         4.83398438e-02,  7.75366892e-03,  6.93203732e+00]),\n",
       " array([-4.81498932e+02,  5.18375206e+01,  3.31722603e+01, ...,\n",
       "         2.39257812e-02,  2.50747363e-02,  1.45807892e+01]),\n",
       " array([-5.01438477e+02,  2.13380585e+01, -2.30302525e+01, ...,\n",
       "         6.15234375e-02,  5.73960391e-03,  1.09927490e+01]),\n",
       " array([-1.60895126e+02, -6.05438709e+00, -4.61493206e+00, ...,\n",
       "         3.90625000e-02,  3.59812452e-02,  1.92177415e+01]),\n",
       " array([-2.36960434e+02,  1.11926399e+02, -4.38269196e+01, ...,\n",
       "         4.83398438e-02,  2.54825383e-02,  1.68367535e+01]),\n",
       " array([-1.75540405e+02,  9.07225037e+00,  5.90643311e+00, ...,\n",
       "         8.64257812e-02,  1.01283136e-02,  1.48559496e+01]),\n",
       " array([-3.57841888e+02,  1.31734772e+02,  2.59560928e+01, ...,\n",
       "         1.58203125e-01,  3.36977754e-02,  2.07809210e+01]),\n",
       " array([-6.22517761e+02,  7.94668198e+01,  6.65882254e+00, ...,\n",
       "         8.30078125e-02,  7.84873979e-03,  1.18145809e+01]),\n",
       " array([-1.46668640e+02,  4.62455330e+01, -1.72000580e+01, ...,\n",
       "         1.15722656e-01,  3.24399280e-02,  1.77741379e+01]),\n",
       " array([-1.04646057e+03, -9.24022102e+00,  3.09161091e+00, ...,\n",
       "         3.71093750e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.35153900e+02,  1.09083237e+02, -8.44167938e+01, ...,\n",
       "         1.41113281e-01,  8.84825650e-03,  2.57641107e+01]),\n",
       " array([-4.86338654e+02,  9.61089325e+01,  3.90719376e+01, ...,\n",
       "         2.20214844e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-6.31871826e+02,  6.42930069e+01,  3.02855530e+01, ...,\n",
       "         5.56640625e-02,  1.98354960e-02,  1.63711712e+01]),\n",
       " array([-2.75148499e+02, -1.35629139e+01,  6.31930466e+01, ...,\n",
       "         5.66406250e-02,  1.20921539e-02,  2.09726050e+01]),\n",
       " array([-2.54746674e+02,  1.85221222e+02, -5.14657211e+01, ...,\n",
       "         1.92871094e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.63216736e+02, -1.22529121e+02, -1.16322117e+01, ...,\n",
       "         9.76562500e-02,  1.20056556e-02,  2.08153799e+01]),\n",
       " array([-4.05362885e+02, -8.84423065e+01,  5.29466553e+01, ...,\n",
       "         5.37109375e-02,  7.57775110e-02,  2.34232277e+01]),\n",
       " array([-5.18281555e+02,  1.57088697e-01,  1.56201243e-01, ...,\n",
       "         5.37109375e-02,  6.18916138e-03,  2.25963026e+01]),\n",
       " array([-1.16678574e+02,  1.93531799e+02, -3.11616364e+01, ...,\n",
       "         6.39648438e-02,  1.11574826e-02,  1.72703341e+01]),\n",
       " array([-5.55144043e+02,  5.88939819e+01,  3.79415207e+01, ...,\n",
       "         1.56250000e-01,  3.66870476e-03,  2.06751674e+01]),\n",
       " array([-3.70764160e+02, -7.45670471e+01,  8.75174255e+01, ...,\n",
       "         2.14843750e-02,  1.30810393e-02,  2.37070084e+01]),\n",
       " array([-4.97326752e+02,  6.38981667e+01,  7.28371143e+00, ...,\n",
       "         4.54101562e-02,  3.16215787e-02,  2.19890937e+01]),\n",
       " array([-4.47396606e+02,  6.81965942e+01,  9.34362507e+00, ...,\n",
       "         2.73437500e-02,  6.12429470e-03,  9.57536176e+00]),\n",
       " array([-1.59999573e+02,  1.35669922e+02,  4.00200996e+01, ...,\n",
       "         1.85546875e-02,  3.81521961e-02,  2.03760266e+01]),\n",
       " array([-4.90982147e+02,  8.14576340e+01,  3.45326424e+00, ...,\n",
       "         4.58984375e-02,  6.91572570e-03,  1.55930981e+01]),\n",
       " array([-2.05546707e+02,  1.21192978e+02,  5.64038897e+00, ...,\n",
       "         3.17382812e-02,  1.28551353e-02,  1.94679275e+01]),\n",
       " array([-3.06455505e+02,  9.06906433e+01,  3.60120010e+01, ...,\n",
       "         7.86132812e-02,  1.68242968e-02,  1.99241921e+01]),\n",
       " array([-1.06213516e+02,  3.50705414e+01, -6.07137451e+01, ...,\n",
       "         1.28417969e-01,  1.21342699e-02,  1.27336413e+01]),\n",
       " array([-1.71992996e+02,  1.00401398e+02, -4.62254095e+00, ...,\n",
       "         8.93554688e-02,  2.55578331e-02,  1.71079680e+01]),\n",
       " array([-2.27595886e+02, -3.62930145e+01,  5.68561363e+01, ...,\n",
       "         5.12695312e-02,  1.92915482e-02,  2.10645035e+01]),\n",
       " array([-9.82400894e+01,  1.17585793e+02, -3.03211823e+01, ...,\n",
       "         1.52832031e-01,  1.69220698e-02,  3.20995748e+01]),\n",
       " array([-4.03481934e+02,  6.03854942e+01, -1.10949192e+01, ...,\n",
       "         9.81445312e-02,  1.43998311e-02,  2.29746073e+01]),\n",
       " array([-1.21327332e+02, -3.23004761e+01,  1.38461857e+01, ...,\n",
       "         1.70898438e-02,  6.80716868e-03,  2.09136248e+01]),\n",
       " array([-4.63850403e+02,  1.26521423e+02,  1.40185623e+01, ...,\n",
       "         1.12792969e-01,  1.57076646e-02,  2.90732205e+01]),\n",
       " array([-2.11697205e+02,  1.36767242e+02,  3.28535247e+00, ...,\n",
       "         2.00683594e-01,  2.80725318e-02,  1.90492049e+01]),\n",
       " array([-2.45928574e+02,  6.15074806e+01,  1.67157402e+01, ...,\n",
       "         4.54101562e-02,  1.58915071e-02,  1.95827693e+01]),\n",
       " array([-2.41070709e+02,  1.86001617e+02,  2.28272653e+00, ...,\n",
       "         7.08007812e-02,  1.88305572e-02,  1.40785918e+01]),\n",
       " array([-5.56021240e+02,  5.11222076e+01,  1.81502914e+01, ...,\n",
       "         3.90625000e-02,  2.14912753e-02,  2.18546927e+01]),\n",
       " array([-2.19460968e+02,  1.52956665e+02, -5.05268784e+01, ...,\n",
       "         2.03125000e-01,  1.78972452e-02,  2.45662391e+01]),\n",
       " array([-2.06793686e+02,  1.46977570e+02, -5.12491074e+01, ...,\n",
       "         2.31933594e-01,  5.11087516e-03,  1.66582048e+01]),\n",
       " array([-5.18590759e+02,  2.59347382e+01, -1.10960598e+01, ...,\n",
       "         1.68945312e-01,  9.42218052e-03,  1.74516901e+01]),\n",
       " array([-4.42107880e+02,  6.94863281e+01, -6.60369949e+01, ...,\n",
       "         1.97265625e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-5.51129211e+02,  2.54534607e+01,  5.41103268e+00, ...,\n",
       "         4.39453125e-02,  1.15521700e-02,  2.30227366e+01]),\n",
       " array([-4.04278198e+02, -5.72706842e+00,  5.80310059e+01, ...,\n",
       "         1.92871094e-01,  1.20510546e-02,  2.40707040e+01]),\n",
       " array([-5.30904053e+02,  6.67358704e+01,  5.08392792e+01, ...,\n",
       "         5.71289062e-02,  1.17238793e-02,  2.30981231e+01]),\n",
       " array([-4.42519623e+02,  6.16331558e+01, -1.95614958e+00, ...,\n",
       "         1.20605469e-01,  1.28249129e-02,  1.60422385e+01]),\n",
       " array([-4.16900909e+02, -7.52238617e+01,  3.23403702e+01, ...,\n",
       "         2.05566406e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.71218460e+02,  1.51860291e+02,  8.04138756e+00, ...,\n",
       "         3.41308594e-01,  1.20461832e-02,  1.71608359e+01]),\n",
       " array([-1.99772095e+02,  1.59086365e+02,  5.02966995e+01, ...,\n",
       "         3.29101562e-01,  1.30870456e-02,  1.21830262e+01]),\n",
       " array([-2.49264023e+02,  7.37733078e+01, -3.79405928e+00, ...,\n",
       "         2.70019531e-01,  1.36540316e-02,  3.65840077e+01]),\n",
       " array([-5.67102661e+02,  2.28438282e+01,  1.49786110e+01, ...,\n",
       "         3.62304688e-01,  6.68571035e-03,  2.44041443e+01]),\n",
       " array([-1.87095169e+02,  1.66851212e+02, -4.80920563e+01, ...,\n",
       "         3.90625000e-02,  7.76614513e-03,  1.54199883e+01]),\n",
       " array([-5.03776245e+02,  7.04664154e+01,  2.48398590e+01, ...,\n",
       "         5.32226562e-02,  1.34886392e-02,  2.79298335e+01]),\n",
       " array([-6.01946411e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         2.68554688e-02,  1.10676236e-02,  1.35103777e+01]),\n",
       " array([-3.32897675e+02,  1.63071823e+02,  2.73609772e+01, ...,\n",
       "         1.46484375e-01,  2.55678378e-02,  1.37537897e+01]),\n",
       " array([-2.46335251e+02,  8.35813217e+01, -4.60986671e+01, ...,\n",
       "         1.19140625e-01,  6.16483425e-03,  1.85490519e+01]),\n",
       " array([-5.67089783e+02,  5.85571594e+01,  3.18383694e+01, ...,\n",
       "         1.59667969e-01,  2.88519311e-02,  2.29307801e+01]),\n",
       " array([-3.53944366e+02,  2.55271378e+01, -2.35578365e+01, ...,\n",
       "         1.09863281e-01,  4.15846947e-02,  1.97494149e+01]),\n",
       " array([-3.19293579e+02, -4.50585861e+01,  4.03386002e+01, ...,\n",
       "         1.02539062e-01,  1.14478729e-02,  7.38155544e+00]),\n",
       " array([-4.51028931e+02,  7.53275299e+01,  1.53362064e+01, ...,\n",
       "         5.07812500e-02,  7.61911255e-03,  1.49733424e+01]),\n",
       " array([-4.37354004e+02,  3.53188438e+01, -2.89083633e+01, ...,\n",
       "         1.99707031e-01,  1.08752681e-02,  1.51380837e+01]),\n",
       " array([-3.64957123e+02,  8.08947220e+01,  4.54636307e+01, ...,\n",
       "         5.71289062e-02,  2.40910480e-02,  2.99960762e+01]),\n",
       " array([-5.56859253e+02,  3.32339168e+00,  3.18781948e+00, ...,\n",
       "         8.49609375e-02,  6.38237286e-03,  1.07462473e+01]),\n",
       " array([-5.15864563e+02,  7.57901788e-01,  7.57110834e-01, ...,\n",
       "         3.71093750e-02,  1.14848176e-02,  2.65564620e+01]),\n",
       " array([-5.66029663e+02,  2.49350662e+01,  1.06399431e+01, ...,\n",
       "         9.32617188e-02,  5.00334187e-03,  1.74554363e+01]),\n",
       " array([-4.57044312e+02,  8.09345016e+01,  1.99713364e+01, ...,\n",
       "         3.61328125e-02,  7.16152828e-03,  2.25248113e+01]),\n",
       " array([-5.73862549e+02,  5.23106613e+01,  2.08602314e+01, ...,\n",
       "         9.86328125e-02,  3.10509403e-02,  1.78028822e+01]),\n",
       " array([-5.05912537e+02,  4.96037865e+01,  4.21292648e+01, ...,\n",
       "         2.78320312e-02,  1.33776177e-02,  2.01263681e+01]),\n",
       " array([-2.64880188e+02,  1.33915329e+02,  2.46099548e+01, ...,\n",
       "         2.44140625e-02,  3.97342174e-03,  1.46759167e+01]),\n",
       " array([-4.09345795e+02,  1.23341782e+02,  3.00149040e+01, ...,\n",
       "         4.15039062e-02,  3.71390845e-02,  1.18637651e+01]),\n",
       " array([-4.12585052e+02,  1.08121849e+02,  9.85652208e-02, ...,\n",
       "         3.07128906e-01,  1.45666083e-02,  2.39665881e+01]),\n",
       " array([-2.83054169e+02,  1.30399597e+02, -2.08953476e+01, ...,\n",
       "         1.90429688e-02,  7.51288712e-03,  1.62575066e+01]),\n",
       " array([-4.33488647e+02,  2.16491814e+01,  6.39835014e+01, ...,\n",
       "         9.57031250e-02,  1.93445898e-02,  1.68529600e+01]),\n",
       " array([-4.37531982e+02,  1.27364075e+02,  1.05103855e+01, ...,\n",
       "         7.51953125e-02,  1.82594759e-02,  2.01462939e+01]),\n",
       " array([-3.76406097e+02,  5.32220383e+01,  6.19753218e+00, ...,\n",
       "         2.06054688e-01,  1.18235570e-02,  2.45226920e+01]),\n",
       " array([-6.51485962e+02,  1.13459749e+01,  7.36655045e+00, ...,\n",
       "         2.09960938e-01,  1.26330341e-02,  1.90221772e+01]),\n",
       " array([-6.18451782e+02,  2.56629868e+01,  2.01937294e+01, ...,\n",
       "         4.98046875e-02,  2.27146153e-02,  1.28983438e+01]),\n",
       " array([-5.52083435e+02,  1.42656541e+00,  1.27167332e+00, ...,\n",
       "         2.14843750e-02,  3.53535982e-02,  2.35839903e+01]),\n",
       " array([-3.29374603e+02,  1.19276581e+01,  5.22964363e+01, ...,\n",
       "         4.15039062e-02,  5.69743662e-02,  2.16106713e+01]),\n",
       " array([-2.61552826e+02,  1.09238930e+01,  3.94502258e+01, ...,\n",
       "         7.86132812e-02,  8.54645928e-03,  1.74690440e+01]),\n",
       " array([-3.62271667e+02,  8.09217987e+01, -1.64654541e+01, ...,\n",
       "         1.44042969e-01,  6.37584147e-03,  1.33194938e+01]),\n",
       " array([-1.66305679e+02,  9.69415588e+01, -4.83526230e+01, ...,\n",
       "         7.81250000e-02,  4.90552418e-03,  1.98807791e+01]),\n",
       " array([-2.32269119e+02,  6.18016052e+01, -9.88360901e+01, ...,\n",
       "         4.19921875e-02,  9.32355058e-03,  1.73657641e+01]),\n",
       " array([-1.13456024e+02,  1.21900345e+02, -2.32602043e+01, ...,\n",
       "         1.41601562e-01,  7.41473064e-03,  1.55174926e+01]),\n",
       " array([-4.75333618e+02, -3.56628113e+01,  4.79075241e+00, ...,\n",
       "         1.52343750e-01,  1.39003635e-02,  1.51895136e+01]),\n",
       " array([-2.25744736e+02,  1.45648834e+02, -4.25497971e+01, ...,\n",
       "         9.52148438e-02,  9.73773638e-03,  1.46973759e+01]),\n",
       " array([-4.93800964e+02,  4.43807526e+01, -3.39042549e+01, ...,\n",
       "         1.09375000e-01,  1.37624371e-02,  1.08587593e+01]),\n",
       " array([-5.51398254e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         8.34960938e-02,  1.46518748e-02,  1.11906938e+01]),\n",
       " array([-6.66152466e+02,  3.82381363e+01,  2.23984599e+00, ...,\n",
       "         1.29882812e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.83144363e+02,  9.86274948e+01, -1.78102684e+01, ...,\n",
       "         6.15234375e-02,  2.42924460e-02,  2.46554762e+01]),\n",
       " array([-4.98853394e+02,  5.99491119e+01,  1.28410397e+01, ...,\n",
       "         1.15234375e-01,  1.41006346e-02,  2.13989943e+01]),\n",
       " array([-4.52186066e+02,  9.58674622e+01,  3.26889801e+01, ...,\n",
       "         1.42578125e-01,  1.15668853e-02,  2.33715817e+01]),\n",
       " array([-3.97290070e+02,  1.13187286e+02,  1.03540754e+01, ...,\n",
       "         1.40625000e-01,  1.20111383e-02,  2.37928718e+01]),\n",
       " array([-5.91654846e+02,  2.82570648e+01,  2.32693024e+01, ...,\n",
       "         6.00585938e-02,  2.80784316e-02,  3.24006587e+01]),\n",
       " array([-2.72474762e+02,  1.05728897e+02,  2.37659779e+01, ...,\n",
       "         8.88671875e-02,  7.85423811e-03,  1.69533744e+01]),\n",
       " array([-2.15220612e+02,  7.30075378e+01, -4.15986481e+01, ...,\n",
       "         1.57226562e-01,  4.16357987e-03,  1.74485520e+01]),\n",
       " array([-4.20485382e+02,  7.20468903e+01,  2.59856224e+01, ...,\n",
       "         7.61718750e-02,  6.91355472e-03,  1.32945761e+01]),\n",
       " array([-5.57802856e+02,  5.79803162e+01,  2.42942505e+01, ...,\n",
       "         2.09960938e-02,  4.72588315e-03,  1.62313744e+01]),\n",
       " array([-2.41043289e+02,  1.08620087e+02,  2.06837730e+01, ...,\n",
       "         1.75292969e-01,  1.05578958e-02,  2.26188928e+01]),\n",
       " array([-3.34775818e+02, -1.17780426e+02,  1.54884262e+01, ...,\n",
       "         1.82128906e-01,  6.98388238e-03,  2.69279599e+01]),\n",
       " array([-4.33129211e+02,  7.40053406e+01,  6.76093435e+00, ...,\n",
       "         2.34375000e-02,  1.58031203e-02,  1.56412959e+01]),\n",
       " array([-6.09856567e+02,  4.29559797e-01,  4.29462761e-01, ...,\n",
       "         1.09375000e-01,  1.35304622e-02,  1.13482572e+01]),\n",
       " array([-4.87060303e+02,  7.63543167e+01, -3.42014694e+01, ...,\n",
       "         1.78222656e-01,  1.18736223e-02,  1.16564482e+01]),\n",
       " array([-2.09100540e+02, -5.08572845e+01, -3.72264328e+01, ...,\n",
       "         1.26953125e-02,  9.78159756e-03,  1.56540036e+01]),\n",
       " array([-5.70449402e+02,  6.37798071e+00,  3.56868005e+00, ...,\n",
       "         1.03515625e-01,  1.45341797e-02,  1.70919850e+01]),\n",
       " array([-4.35519196e+02,  1.25096466e+02,  2.11834259e+01, ...,\n",
       "         5.56640625e-02,  3.19794297e-02,  1.00298561e+01]),\n",
       " array([-4.08379852e+02,  9.60482178e+01,  8.88738036e-01, ...,\n",
       "         4.14550781e-01,  1.08497576e-02,  1.06371149e+01]),\n",
       " array([-5.52425171e+02,  3.38591728e+01, -4.63104477e+01, ...,\n",
       "         4.98046875e-02,  6.01508375e-03,  1.32628664e+01]),\n",
       " array([-180.7538147 ,  -73.12454224,   54.00891876, ...,    0.20019531,\n",
       "           0.        ,    0.        ]),\n",
       " array([-1.37369598e+02,  1.27735611e+02, -1.19797821e+01, ...,\n",
       "         6.73828125e-02,  1.05262734e-02,  2.93122768e+01]),\n",
       " array([-1.96356735e+01,  4.27185364e+01, -3.92993240e+01, ...,\n",
       "         1.44042969e-01,  1.74578771e-02,  1.92887366e+01]),\n",
       " array([-6.45160889e+02,  1.86881378e-01,  1.76415339e-01, ...,\n",
       "         8.83789062e-02,  2.17883509e-02,  1.65149748e+01]),\n",
       " array([-6.02341187e+02,  2.53457725e-01,  2.43760675e-01, ...,\n",
       "         4.73632812e-02,  3.24996671e-02,  1.59554660e+01]),\n",
       " array([-3.61518280e+02,  1.25020447e+02,  7.78539848e+00, ...,\n",
       "         1.17675781e-01,  9.22649028e-03,  1.38590753e+01]),\n",
       " array([-5.09189453e+02,  4.47190018e+01,  5.80983829e+00, ...,\n",
       "         9.27734375e-02,  1.47588085e-02,  2.37187445e+01]),\n",
       " array([-3.02990387e+02,  9.45994797e+01,  3.68580551e+01, ...,\n",
       "         3.71093750e-02,  1.13514265e-02,  1.20206669e+01]),\n",
       " array([-350.79516602,  141.26315308,  -11.50137138, ...,    0.41699219,\n",
       "           0.        ,    0.        ]),\n",
       " array([-5.62457092e+02,  2.04723778e+01,  5.75443935e+00, ...,\n",
       "         1.66015625e-01,  8.03988443e-03,  7.44537264e+00]),\n",
       " array([-5.36337891e+02,  1.13155731e+02, -1.38176470e+01, ...,\n",
       "         9.37500000e-02,  1.83432944e-02,  2.63676882e+01]),\n",
       " array([-4.54311340e+02,  1.07808914e+02,  1.92380753e+01, ...,\n",
       "         2.29492188e-02,  3.47807145e-02,  1.50668710e+01]),\n",
       " array([-2.59870575e+02,  3.66950989e+01, -6.20592499e+01, ...,\n",
       "         1.43554688e-01,  7.88977136e-03,  1.50908291e+01]),\n",
       " array([-4.99405945e+02,  9.09940948e+01,  3.32116814e+01, ...,\n",
       "         4.68750000e-02,  1.97645081e-02,  1.28896222e+01]),\n",
       " array([-2.21391922e+02,  7.45063400e+01,  4.19069061e+01, ...,\n",
       "         2.13378906e-01,  1.59014212e-02,  1.02645010e+01]),\n",
       " array([-4.74355011e+02,  1.19084351e+02,  3.33072701e+01, ...,\n",
       "         1.22070312e-01,  2.27366371e-02,  1.89893335e+01]),\n",
       " array([-1.86072205e+02,  9.79691772e+01, -4.31283264e+01, ...,\n",
       "         1.27441406e-01,  1.00811546e-02,  2.21937373e+01]),\n",
       " array([-3.03257660e+02,  1.03885635e+02,  1.78005867e+01, ...,\n",
       "         1.33789062e-01,  1.75913913e-02,  1.75478727e+01]),\n",
       " array([-4.63755524e+02,  8.25391006e+01,  6.20617390e+00, ...,\n",
       "         8.44726562e-02,  2.53472084e-02,  1.89641088e+01]),\n",
       " array([-2.98175446e+02,  9.58635864e+01, -3.47193680e+01, ...,\n",
       "         2.49023438e-02,  1.42095788e-02,  2.08730489e+01]),\n",
       " array([-1.40129318e+02,  1.37307770e+02,  1.14559758e+00, ...,\n",
       "         5.21972656e-01,  1.84298197e-02,  2.31864735e+01]),\n",
       " array([-5.78779663e+02,  2.26512527e+01, -1.30707111e+01, ...,\n",
       "         1.59667969e-01,  2.56785304e-02,  1.63543731e+01]),\n",
       " array([-3.92115143e+02,  1.38120407e+02,  4.56051865e+01, ...,\n",
       "         9.27734375e-02,  1.31120174e-02,  1.19822964e+01]),\n",
       " array([-3.25030640e+02,  8.38824005e+01,  2.85865364e+01, ...,\n",
       "         5.37109375e-02,  5.27712121e-02,  1.76251084e+01]),\n",
       " array([-4.93042480e+02,  1.29971638e+01, -2.01060143e+01, ...,\n",
       "         1.97265625e-01,  5.37896485e-03,  1.94265246e+01]),\n",
       " array([-3.72765472e+02,  1.73876968e+02,  3.63088608e+01, ...,\n",
       "         4.68750000e-02,  4.51447073e-02,  1.91923365e+01]),\n",
       " array([-5.12136841e+02,  1.49543877e+01, -1.09952374e+01, ...,\n",
       "         4.62890625e-01,  1.01120007e-02,  1.70086861e+01]),\n",
       " array([-1.64811462e+02,  1.28767746e+02, -1.31982145e+01, ...,\n",
       "         1.11328125e-01,  1.37515440e-02,  9.41953734e+00]),\n",
       " array([-5.26420593e+02,  6.39133606e+01,  4.47436676e+01, ...,\n",
       "         6.64062500e-02,  2.06486153e-02,  2.12637395e+01]),\n",
       " array([-2.66823914e+02,  1.36339050e+02,  4.18528748e+01, ...,\n",
       "         6.49414062e-02,  3.71584035e-02,  1.91573694e+01]),\n",
       " array([-5.73411987e+02,  3.25968895e+01, -8.40904176e-01, ...,\n",
       "         1.67480469e-01,  1.10839241e-02,  1.74722657e+01]),\n",
       " array([-5.33083374e+02,  2.75344734e+01, -1.31371379e+00, ...,\n",
       "         1.43066406e-01,  6.15237112e-03,  1.61210552e+01]),\n",
       " array([-2.36539215e+02,  1.29204041e+02,  1.21880741e+01, ...,\n",
       "         5.51757812e-02,  2.45329765e-02,  8.34206194e+00]),\n",
       " array([-2.40336075e+02,  1.58020844e+02,  1.22391195e+01, ...,\n",
       "         4.19921875e-02,  1.23136110e-02,  1.71407178e+01]),\n",
       " array([-4.70716034e+02,  5.64583282e+01,  8.35870361e+00, ...,\n",
       "         9.17968750e-02,  1.38434718e-02,  9.90841240e+00]),\n",
       " array([-1.86637726e+02,  1.21209320e+02, -1.97168732e+01, ...,\n",
       "         1.94335938e-01,  1.07682841e-02,  1.98427096e+01]),\n",
       " array([-2.31207901e+02,  4.60770187e+01,  2.59869194e+01, ...,\n",
       "         1.90429688e-02,  2.86809892e-02,  1.92325294e+01]),\n",
       " array([-4.32688019e+02, -4.58390121e+01, -4.83403549e+01, ...,\n",
       "         1.43066406e-01,  2.01863719e-02,  3.70675027e+01]),\n",
       " array([-6.08116333e+02,  4.14293365e+01, -2.95076942e+01, ...,\n",
       "         1.94335938e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-3.53272217e+02, -1.11338470e+02,  2.14203262e+01, ...,\n",
       "         3.27148438e-02,  1.63471235e-02,  1.17057070e+01]),\n",
       " array([-2.67465118e+02,  1.58253662e+02, -3.45988655e+01, ...,\n",
       "         7.27539062e-02,  2.96381015e-02,  1.93010047e+01]),\n",
       " array([-2.55564758e+02, -1.08043243e+02, -1.86732941e+01, ...,\n",
       "         6.29882812e-02,  4.47153815e-03,  1.38385817e+01]),\n",
       " array([-2.67433868e+02,  6.08141365e+01,  3.21250839e+01, ...,\n",
       "         5.61523438e-02,  1.40026910e-02,  1.22200571e+01]),\n",
       " array([-3.78597839e+02,  1.32004608e+02,  2.51082096e+01, ...,\n",
       "         1.91406250e-01,  1.77125390e-02,  2.70544767e+01]),\n",
       " array([-4.82179352e+02,  9.70447845e+01,  3.26721992e+01, ...,\n",
       "         9.08203125e-02,  2.80185225e-02,  1.83552101e+01]),\n",
       " array([-5.23281372e+02,  2.04037552e+01,  8.54362774e+00, ...,\n",
       "         6.44531250e-02,  2.69928418e-02,  1.72265634e+01]),\n",
       " array([-5.15672791e+02,  1.66487656e+01, -9.58997536e+00, ...,\n",
       "         4.00390625e-02,  8.05418074e-03,  1.76019490e+01]),\n",
       " array([-2.96149719e+02, -4.50386810e+01,  6.71750927e+00, ...,\n",
       "         2.44140625e-02,  1.39474270e-02,  1.48010164e+01]),\n",
       " array([-4.55871185e+02, -8.85215607e+01,  5.61349983e+01, ...,\n",
       "         5.17578125e-02,  1.78493219e-02,  4.07180488e+01]),\n",
       " array([-1.11053162e+02,  1.31369781e+02, -1.26065880e+02, ...,\n",
       "         1.46972656e-01,  1.07464701e-02,  1.52813151e+01]),\n",
       " array([-5.72916016e+02,  4.04235649e+01,  8.69218540e+00, ...,\n",
       "         2.73437500e-01,  1.35028425e-02,  2.58717269e+01]),\n",
       " array([-3.95208801e+02, -2.86014462e+01, -6.15919456e+01, ...,\n",
       "         1.84570312e-01,  1.18633970e-02,  2.02649519e+01]),\n",
       " array([-2.73169830e+02,  1.35894562e+02, -5.56095047e+01, ...,\n",
       "         1.10839844e-01,  2.19278341e-02,  1.72418892e+01]),\n",
       " array([-2.12704239e+02,  6.85343323e+01, -5.09853897e+01, ...,\n",
       "         2.24121094e-01,  3.10055022e-02,  2.29730025e+01]),\n",
       " array([-1.43111374e+02,  1.11443451e+02, -3.61956940e+01, ...,\n",
       "         1.04003906e-01,  1.98949895e-02,  1.56469464e+01]),\n",
       " array([-3.36885376e+02,  1.44433624e+02,  4.58565750e+01, ...,\n",
       "         1.73339844e-01,  9.23787203e-03,  1.92279398e+01]),\n",
       " array([-3.45037598e+02,  1.32183685e+02, -2.32047863e+01, ...,\n",
       "         2.83203125e-02,  6.71864147e-02,  2.49207109e+01]),\n",
       " array([-2.26039108e+02,  1.92853394e+01,  9.76830006e+00, ...,\n",
       "         6.64062500e-02,  2.96347866e-02,  1.23868614e+01]),\n",
       " array([-1.81217819e+02,  1.05932770e+02,  2.14011898e+01, ...,\n",
       "         4.29687500e-02,  1.05280025e-02,  3.38675678e+01]),\n",
       " array([-4.16471985e+02,  8.33240662e+01,  5.19396896e+01, ...,\n",
       "         4.47265625e-01,  9.51330515e-03,  1.58682704e+01]),\n",
       " array([-1.64680969e+02,  1.47047424e+02, -3.96754913e+01, ...,\n",
       "         3.90625000e-02,  2.25038092e-02,  2.02400029e+01]),\n",
       " array([-2.32794022e+02,  1.55046783e+02, -3.76810608e+01, ...,\n",
       "         2.11914062e-01,  9.13690905e-03,  9.69680324e+00]),\n",
       " array([-2.78807098e+02,  1.15810905e+02, -6.92641449e+01, ...,\n",
       "         1.52832031e-01,  4.75592735e-03,  2.09935442e+01]),\n",
       " array([-4.33670074e+02,  1.51753036e+02,  2.24217873e+01, ...,\n",
       "         9.86328125e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.32137268e+02,  1.41637939e+02, -7.34469757e+01, ...,\n",
       "         7.76367188e-02,  2.58914883e-02,  2.28670821e+01]),\n",
       " array([-5.19844116e+02, -8.83029521e-01,  1.78696671e+01, ...,\n",
       "         8.59375000e-02,  9.79867225e-03,  2.50405461e+01]),\n",
       " array([-5.63035767e+02,  6.82601547e+01,  2.67321091e+01, ...,\n",
       "         4.15039062e-02,  8.38858340e-03,  1.19691052e+01]),\n",
       " array([-3.03894165e+02,  1.33122955e+02,  3.50332618e+00, ...,\n",
       "         6.88476562e-02,  3.26856925e-02,  1.95788726e+01]),\n",
       " array([-4.28657318e+02,  3.90602379e+01, -1.30223064e+01, ...,\n",
       "         6.29882812e-02,  6.20014299e-03,  6.34774491e+00]),\n",
       " array([-4.89546844e+02,  8.40983810e+01,  3.74581451e+01, ...,\n",
       "         2.24609375e-02,  4.10163450e-02,  1.95242181e+01]),\n",
       " array([-5.21060791e+02,  3.79043198e+01,  3.71167588e+00, ...,\n",
       "         1.91406250e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.05132431e+02, -6.96564674e+00, -1.29804792e+01, ...,\n",
       "         3.23242188e-01,  2.33163527e-02,  2.38821417e+01]),\n",
       " array([-6.30007263e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         2.07519531e-01,  4.15163649e-03,  4.88557257e+00]),\n",
       " array([-2.56433502e+02,  7.57698059e+01, -2.57635384e+01, ...,\n",
       "         1.71875000e-01,  9.59177742e-03,  2.46532932e+01]),\n",
       " array([-3.44062897e+02,  1.60540924e+02, -2.09066010e+01, ...,\n",
       "         7.61718750e-02,  3.06021707e-02,  1.33907929e+01]),\n",
       " array([-1.52927719e+02,  3.75533485e+01, -3.53708420e+01, ...,\n",
       "         2.36328125e-01,  7.73262210e-03,  1.79717675e+01]),\n",
       " array([-2.02464508e+02, -2.41891742e+00,  3.19172840e+01, ...,\n",
       "         1.07910156e-01,  8.41900093e-03,  1.65860161e+01]),\n",
       " array([-5.73804871e+02,  7.65752673e-01,  6.39070570e-01, ...,\n",
       "         5.22460938e-02,  7.58079760e-03,  2.20303014e+01]),\n",
       " array([-1.57849808e+02,  8.65922699e+01, -5.53293419e+01, ...,\n",
       "         2.06054688e-01,  9.57250024e-03,  2.59253323e+01]),\n",
       " array([-5.59172974e+02,  1.98069885e-01,  1.95876092e-01, ...,\n",
       "         1.52832031e-01,  2.15652056e-02,  2.14848295e+01]),\n",
       " array([-3.38438225e+01,  6.46080093e+01, -5.44160118e+01, ...,\n",
       "         5.02929688e-02,  2.92551356e-02,  2.09273875e+01]),\n",
       " array([-2.80097504e+02,  1.32837189e+02, -5.59289122e+00, ...,\n",
       "         2.88085938e-02,  1.91669300e-02,  1.31895214e+01]),\n",
       " array([-1.45764374e+02,  9.83122559e+01, -6.72331390e+01, ...,\n",
       "         3.95507812e-02,  6.87149688e-03,  1.80566490e+01]),\n",
       " array([-7.71489029e+01,  1.05026184e+02, -3.36472473e+01, ...,\n",
       "         1.79199219e-01,  8.35759106e-03,  6.75734356e+00]),\n",
       " array([-4.79705566e+02,  1.80877614e+00, -8.42457104e+00, ...,\n",
       "         9.32617188e-02,  6.21447435e-03,  1.05655380e+01]),\n",
       " array([-1.96571457e+02,  1.85820801e+02, -5.11765022e+01, ...,\n",
       "         1.44042969e-01,  1.65841468e-02,  1.82293206e+01]),\n",
       " array([-3.72384033e+02,  1.88181396e+02, -5.29750204e+00, ...,\n",
       "         8.44726562e-02,  3.63965980e-02,  1.57153323e+01]),\n",
       " array([-4.20628845e+02,  1.83425781e+02,  3.06170788e+01, ...,\n",
       "         4.98046875e-02,  2.83563343e-02,  1.52847081e+01]),\n",
       " array([-2.59010315e+02, -5.31677017e+01,  2.19126511e+01, ...,\n",
       "         2.06054688e-01,  7.14438846e-03,  2.06130296e+01]),\n",
       " array([-2.91063171e+02,  1.39015930e+02,  3.82291069e+01, ...,\n",
       "         2.58789062e-02,  2.13687048e-02,  2.66109824e+01]),\n",
       " array([-2.88949585e+02,  7.49229279e+01,  1.32538915e+00, ...,\n",
       "         1.79687500e-01,  1.51546028e-02,  2.69285053e+01]),\n",
       " array([-4.81794525e+02,  5.32113419e+01, -1.44861090e+00, ...,\n",
       "         6.98242188e-02,  1.44817707e-02,  1.92107275e+01]),\n",
       " array([-1.39360260e+02,  7.65566940e+01, -2.51061249e+01, ...,\n",
       "         1.43554688e-01,  3.84067339e-03,  1.42653763e+01]),\n",
       " array([-2.68072876e+02,  1.38565887e+02,  4.26296692e+01, ...,\n",
       "         2.29492188e-02,  3.64037753e-02,  2.48146757e+01]),\n",
       " array([-5.60746338e+02,  1.28944874e+01, -1.67901287e+01, ...,\n",
       "         1.07421875e-01,  2.82519543e-02,  1.69467270e+01]),\n",
       " array([-5.55242615e+02,  6.45229340e+00, -8.33187866e+00, ...,\n",
       "         7.03125000e-02,  9.95689675e-03,  1.79031044e+01]),\n",
       " array([-5.84436768e+02,  1.67022491e+00,  9.65366006e-01, ...,\n",
       "         1.67968750e-01,  8.19874437e-03,  3.29171121e+01]),\n",
       " array([-5.20592957e+02,  1.10020819e+01,  3.13098145e+00, ...,\n",
       "         2.45605469e-01,  9.42035485e-03,  1.24758862e+01]),\n",
       " array([-3.88429047e+02,  1.08241737e+02,  2.50067139e+01, ...,\n",
       "         2.03613281e-01,  1.91733499e-02,  1.80133060e+01]),\n",
       " array([-3.93982452e+02,  1.16064247e+02,  2.84987030e+01, ...,\n",
       "         3.16406250e-01,  9.40893523e-03,  1.66220427e+01]),\n",
       " array([-6.69202637e+02,  5.19417191e+00,  5.02580070e+00, ...,\n",
       "         1.70410156e-01,  1.65510711e-02,  1.13383286e+01]),\n",
       " array([-1.56692322e+02,  1.45313538e+02,  1.15073996e+01, ...,\n",
       "         8.69140625e-02,  8.48122722e-03,  1.80455685e+01]),\n",
       " array([-5.06592072e+02,  8.84945831e+01,  1.48208065e+01, ...,\n",
       "         4.88281250e-02,  2.15994135e-02,  2.20024213e+01]),\n",
       " array([-5.09116020e+01,  8.59289169e+01, -6.67930298e+01, ...,\n",
       "         1.55761719e-01,  1.28609768e-02,  1.89110488e+01]),\n",
       " array([-7.03488159e+01,  1.05020035e+02, -5.89914169e+01, ...,\n",
       "         2.00195312e-02,  7.85589927e-03,  1.68503329e+01]),\n",
       " array([-3.09732330e+02,  6.15142345e+00,  1.47047234e+01, ...,\n",
       "         1.12792969e-01,  7.59921503e-03,  1.15079977e+01]),\n",
       " array([-2.06422546e+02,  1.35526398e+02,  1.05108109e+01, ...,\n",
       "         8.39843750e-02,  1.63592818e-02,  1.49334133e+01]),\n",
       " array([-1.92723541e+02,  9.99217453e+01, -4.37658005e+01, ...,\n",
       "         4.18945312e-01,  1.31404047e-02,  1.98322371e+01]),\n",
       " array([-6.00074036e+02,  7.00413513e+01,  1.35673351e+01, ...,\n",
       "         9.57031250e-02,  4.08451081e-02,  1.66105583e+01]),\n",
       " array([-6.16755493e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         1.95800781e-01,  2.70093926e-02,  1.49111032e+01]),\n",
       " array([-2.96692474e+02,  1.09102112e+02,  1.22485495e+01, ...,\n",
       "         6.59179688e-02,  2.89291381e-02,  1.21437222e+01]),\n",
       " array([-1.99721405e+02,  1.22908493e+02, -2.55612793e+01, ...,\n",
       "         6.98242188e-02,  1.16370665e-02,  1.27712891e+01]),\n",
       " array([-2.41391159e+02,  1.22181870e+02, -2.77171478e+01, ...,\n",
       "         2.19238281e-01,  1.12269169e-02,  2.39140227e+01]),\n",
       " array([-8.78245163e+01,  1.15468323e+02,  1.05625105e+01, ...,\n",
       "         8.78906250e-03,  6.68000066e-03,  9.18719992e+00]),\n",
       " array([-5.31428894e+02,  6.84101410e+01,  4.97559128e+01, ...,\n",
       "         1.61132812e-02,  1.37366317e-02,  1.84562936e+01]),\n",
       " array([-3.50428497e+02,  1.86273682e+02, -3.16069298e+01, ...,\n",
       "         6.64062500e-02,  2.25236803e-02,  1.86273441e+01]),\n",
       " array([-1.95470993e+02,  1.01018700e+02, -6.02114182e+01, ...,\n",
       "         1.77734375e-01,  1.20478012e-02,  9.75006595e+00]),\n",
       " array([-1.96563202e+02,  1.33583832e+02,  3.79789619e+01, ...,\n",
       "         5.51757812e-02,  2.94673305e-02,  2.19671652e+01]),\n",
       " array([-1.96765274e+02,  8.53479309e+01,  4.51064110e+00, ...,\n",
       "         5.85937500e-02,  1.59512281e-02,  2.37692147e+01]),\n",
       " array([-6.50508179e+02,  7.38375366e-01,  7.38208592e-01, ...,\n",
       "         2.57324219e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-5.77710693e+02, -9.57204437e+00, -2.24575100e+01, ...,\n",
       "         1.57226562e-01,  2.56693999e-02,  2.45262936e+01]),\n",
       " array([-6.01632751e+02,  2.57108727e+01,  6.28618813e+00, ...,\n",
       "         5.95703125e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-5.34252380e+02,  9.08511639e+00, -5.88142014e+00, ...,\n",
       "         1.57714844e-01,  1.29677091e-02,  2.37630546e+01]),\n",
       " array([-2.80872345e+02,  5.26480408e+01,  2.13553448e+01, ...,\n",
       "         1.15234375e-01,  1.33565870e-02,  1.76927641e+01]),\n",
       " array([-1.51610794e+02,  1.16848618e+02,  8.89599037e+00, ...,\n",
       "         8.64257812e-02,  5.97038063e-03,  2.44748428e+01]),\n",
       " array([-6.22413696e+02,  3.37166977e+01, -5.46768112e+01, ...,\n",
       "         2.37792969e-01,  5.04019092e-03,  1.24231897e+01]),\n",
       " array([-4.80118591e+02,  1.16896225e+02,  2.95886192e+01, ...,\n",
       "         2.31445312e-01,  1.18467216e-02,  1.97451800e+01]),\n",
       " array([-9.14123917e+01,  5.62056198e+01, -4.73899803e+01, ...,\n",
       "         3.38378906e-01,  1.26932828e-02,  1.26293942e+01]),\n",
       " array([-2.69496033e+02,  1.66862579e+02, -4.79326820e+00, ...,\n",
       "         2.44140625e-02,  2.19698671e-02,  1.39956355e+01]),\n",
       " array([-3.11166931e+02,  1.18706650e+02, -4.59724808e+01, ...,\n",
       "         7.56835938e-02,  1.35336329e-02,  1.73807397e+01]),\n",
       " array([-2.62648132e+02,  1.42858459e+02,  1.87117825e+01, ...,\n",
       "         3.80859375e-02,  6.80252779e-02,  7.70444050e+00]),\n",
       " array([-3.21927277e+02,  9.04028854e+01,  1.49125156e+01, ...,\n",
       "         4.45312500e-01,  7.76387554e-03,  2.07044795e+01]),\n",
       " array([-3.17675049e+02,  1.48081757e+02, -1.98052025e+01, ...,\n",
       "         9.81445312e-02,  3.70493386e-02,  1.25861615e+01]),\n",
       " array([-5.03602692e+02,  1.16691513e+02,  3.59713974e+01, ...,\n",
       "         4.24804688e-02,  1.32310907e-02,  1.37364566e+01]),\n",
       " array([-1.94730484e+02,  7.51663208e+01, -3.75858650e+01, ...,\n",
       "         1.96289062e-01,  1.42314035e-02,  2.21336231e+01]),\n",
       " array([-6.64116028e+02,  1.04172397e+00,  1.04148865e+00, ...,\n",
       "         6.54296875e-02,  4.22053921e-02,  1.51978940e+01]),\n",
       " array([-2.59788818e+02,  1.41470673e+02, -8.48353767e+00, ...,\n",
       "         3.07617188e-01,  1.72122519e-02,  1.92492992e+01]),\n",
       " array([-5.52620911e+02,  7.02952194e+01,  2.89103203e+01, ...,\n",
       "         7.27539062e-02,  1.74729905e-02,  2.25540638e+01]),\n",
       " array([-3.40556259e+01,  7.74960327e+01, -2.63690586e+01, ...,\n",
       "         9.32617188e-02,  1.40204868e-02,  1.43197805e+01]),\n",
       " array([-1.34197937e+02,  2.96048851e+01,  3.12814522e+01, ...,\n",
       "         2.88085938e-02,  1.75625760e-02,  1.35027215e+01]),\n",
       " array([-1.46647827e+02,  9.51996307e+01, -5.13064194e+01, ...,\n",
       "         2.45117188e-01,  1.47976130e-02,  2.67254889e+01]),\n",
       " array([-1.82419525e+02,  9.70505981e+01, -1.31479347e+00, ...,\n",
       "         1.27441406e-01,  8.36198651e-03,  1.71161711e+01]),\n",
       " array([-4.62928772e+02,  8.78745728e+01,  3.06499634e+01, ...,\n",
       "         8.05664062e-02,  2.97472892e-02,  3.03082407e+01]),\n",
       " array([-5.60935303e+02,  1.68523483e+01,  9.72324944e+00, ...,\n",
       "         2.66601562e-01,  1.13891892e-02,  1.19829327e+01]),\n",
       " array([-4.19242493e+02, -4.62947960e+01, -3.73777580e+01, ...,\n",
       "         1.25976562e-01,  3.86904800e-03,  2.22666293e+01]),\n",
       " array([-2.12549118e+02,  3.50523872e+01,  4.26033325e+01, ...,\n",
       "         2.15820312e-01,  6.42745327e-03,  1.57366842e+01]),\n",
       " array([-1.78804184e+02,  1.18617355e+02,  2.61439285e+01, ...,\n",
       "         3.80859375e-02,  1.07941704e-02,  2.07066715e+01]),\n",
       " array([-5.08312103e+02,  1.99841404e+00, -7.12185478e+00, ...,\n",
       "         8.20312500e-02,  5.85917788e-03,  9.85468030e+00]),\n",
       " array([-2.91991150e+02,  1.45479767e+02, -1.59475327e+01, ...,\n",
       "         1.95312500e-02,  2.22868230e-02,  2.00631514e+01]),\n",
       " array([-3.58544800e+02, -1.09449425e+01,  7.30311203e+01, ...,\n",
       "         1.92871094e-01,  6.97176012e-03,  1.79205820e+01]),\n",
       " array([-1.69751938e+02,  5.69281311e+01, -5.69920349e+01, ...,\n",
       "         1.15234375e-01,  1.33159092e-02,  2.80269891e+01]),\n",
       " array([-6.46812439e+02,  2.00656776e+01,  1.25332918e+01, ...,\n",
       "         2.06054688e-01,  6.17532943e-03,  1.90210789e+01]),\n",
       " array([-4.63209564e+02,  7.87561951e+01,  3.37927704e+01, ...,\n",
       "         3.46679688e-02,  1.64033956e-02,  1.34624735e+01]),\n",
       " array([-5.74284180e+02,  9.52215004e+00, -1.25257626e+01, ...,\n",
       "         3.27148438e-02,  2.35676225e-02,  1.66766629e+01]),\n",
       " array([-4.96479767e+02,  4.91280823e+01, -1.63167953e+00, ...,\n",
       "         1.30859375e-01,  3.91118145e-02,  1.75709665e+01]),\n",
       " array([-3.28535553e+02,  1.65412567e+02, -2.18792610e+01, ...,\n",
       "         4.98046875e-02,  2.44430712e-02,  2.68374920e+01]),\n",
       " array([-4.47595001e+02,  4.63748169e+01, -4.12646332e+01, ...,\n",
       "         2.78320312e-01,  9.12270144e-03,  1.64063454e+01]),\n",
       " array([-6.94423676e+01,  8.75565796e+01, -4.45103226e+01, ...,\n",
       "         1.04003906e-01,  9.94852595e-03,  2.14622483e+01]),\n",
       " array([-2.91948242e+02,  1.37201874e+02,  3.73810959e+01, ...,\n",
       "         4.63867188e-02,  1.82940468e-02,  1.78044602e+01]),\n",
       " array([-5.42803955e+02,  2.17837286e+00,  2.14241982e+00, ...,\n",
       "         8.93554688e-02,  1.00579689e-02,  2.21960589e+01]),\n",
       " array([-4.06221802e+02, -8.79729691e+01,  5.57581139e+01, ...,\n",
       "         1.34277344e-01,  1.30278158e-02,  1.46749973e+01]),\n",
       " array([-2.49692856e+02,  1.68747742e+02,  1.18178844e+01, ...,\n",
       "         1.81640625e-01,  2.67666599e-02,  2.49773517e+01]),\n",
       " array([-5.30198853e+02,  2.33499355e+01, -2.21669102e+00, ...,\n",
       "         1.13281250e-01,  2.09547035e-03,  5.23762032e+00]),\n",
       " array([-3.79508392e+02,  9.95815353e+01,  9.93076992e+00, ...,\n",
       "         6.98242188e-02,  7.61536492e-03,  2.03375474e+01]),\n",
       " array([-3.58485718e+02,  2.89809952e+01, -7.37369843e+01, ...,\n",
       "         3.27148438e-01,  7.07232458e-03,  1.78804085e+01]),\n",
       " array([-3.08017517e+02,  1.70477142e+02, -2.91632919e+01, ...,\n",
       "         2.80273438e-01,  1.26713002e-02,  1.24140918e+01]),\n",
       " array([-1.74166565e+02,  5.98616028e+01, -6.91045685e+01, ...,\n",
       "         1.22070312e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.30190140e+02,  1.00519165e+02, -1.84705658e+01, ...,\n",
       "         5.85937500e-02,  7.87829457e-03,  1.53331891e+01]),\n",
       " array([-5.76175171e+02,  1.74391102e-02,  1.19388020e-02, ...,\n",
       "         3.90625000e-02,  1.03022625e-02,  1.84861138e+01]),\n",
       " array([-1.85871002e+02,  5.68921852e+01, -3.75672102e+00, ...,\n",
       "         1.82617188e-01,  1.42926316e-02,  1.73156172e+01]),\n",
       " array([-5.12185547e+02,  8.81510735e+00, -5.08725643e-01, ...,\n",
       "         1.63574219e-01,  5.81779946e-03,  9.31286886e+00]),\n",
       " array([-1.52931015e+02,  7.15571899e+01,  1.57454815e+01, ...,\n",
       "         2.50000000e-01,  1.04591623e-02,  1.88706130e+01]),\n",
       " array([-8.28376389e+01,  8.02172699e+01, -3.91187515e+01, ...,\n",
       "         4.49218750e-02,  1.76203148e-02,  1.60910949e+01]),\n",
       " array([-3.17086395e+02,  6.96591873e+01,  4.73762665e+01, ...,\n",
       "         5.41992188e-02,  6.05400426e-03,  1.77255973e+01]),\n",
       " array([-2.18663742e+02,  2.98462334e+01, -3.36575508e+01, ...,\n",
       "         3.32031250e-02,  1.61849497e-02,  2.30022117e+01]),\n",
       " array([-5.66965637e+02,  8.00449371e+01,  3.25887108e+01, ...,\n",
       "         3.12500000e-02,  6.66248375e-03,  1.99475393e+01]),\n",
       " array([-2.70532074e+02,  1.62227097e+01,  3.28284378e+01, ...,\n",
       "         3.03222656e-01,  2.14764305e-02,  1.31191567e+01]),\n",
       " array([-4.36913879e+02,  1.64378189e+02,  3.20734329e+01, ...,\n",
       "         2.83203125e-02,  1.78920421e-02,  1.61483631e+01]),\n",
       " array([-2.59905487e+02,  1.41770981e+02,  1.63049603e+01, ...,\n",
       "         7.81250000e-02,  1.46706247e-02,  2.32666984e+01]),\n",
       " array([-5.27186157e+02,  1.85851269e+01,  1.59777012e+01, ...,\n",
       "         7.56835938e-02,  6.86884598e-03,  2.90532023e+01]),\n",
       " array([-5.69932678e+02,  1.54607635e+02,  6.15080490e+01, ...,\n",
       "         8.74023438e-02,  2.90473250e-02,  2.15999112e+01]),\n",
       " array([-1.33757172e+02,  8.42525406e+01, -2.02255592e+01, ...,\n",
       "         8.44726562e-02,  2.02062041e-02,  1.50234029e+01]),\n",
       " array([-5.29477417e+02,  7.33363092e-01,  7.32219338e-01, ...,\n",
       "         1.67480469e-01,  8.04600561e-03,  1.39906257e+01]),\n",
       " array([-5.47036865e+02,  1.03260231e+01,  4.53514624e+00, ...,\n",
       "         9.37500000e-02,  1.49929350e-02,  2.30946526e+01]),\n",
       " array([-2.09446930e+02,  1.38349060e+02, -6.97824478e+00, ...,\n",
       "         3.04687500e-01,  6.27098460e-03,  5.30640632e+00]),\n",
       " array([-6.69036133e+02,  5.91700859e+01, -4.97548828e+01, ...,\n",
       "         1.63574219e-01,  3.11543172e-02,  2.07591414e+01]),\n",
       " array([-1.49172226e+02,  1.34536530e+02, -5.29419937e+01, ...,\n",
       "         6.10351562e-02,  2.34710378e-02,  1.38706192e+01]),\n",
       " array([-5.68744080e+02,  5.03668725e-01,  3.50371420e-01, ...,\n",
       "         3.23730469e-01,  8.73414000e-03,  1.99303061e+01]),\n",
       " array([-2.65554962e+02,  1.03403946e+02, -3.77618332e+01, ...,\n",
       "         2.44628906e-01,  2.57646122e-02,  2.22953230e+01]),\n",
       " array([-4.40752014e+02,  1.11238060e+02,  1.58880806e+01, ...,\n",
       "         7.22656250e-02,  2.65725022e-02,  1.70591488e+01]),\n",
       " array([-5.15211853e+02,  4.57662773e+00, -1.42773676e+01, ...,\n",
       "         1.55761719e-01,  1.04752232e-02,  3.40073973e+01]),\n",
       " array([-1.02099373e+02,  1.90815094e+02, -7.25816345e+01, ...,\n",
       "         2.09960938e-02,  2.64784816e-02,  1.73381507e+01]),\n",
       " array([-5.84157349e+02,  3.70315504e+00,  3.66463208e+00, ...,\n",
       "         1.04980469e-01,  4.90575627e-03,  1.09849922e+01]),\n",
       " array([-4.23650574e+02,  1.16095016e+02,  3.82058640e+01, ...,\n",
       "         2.68554688e-02,  2.21611421e-02,  9.98387262e+00]),\n",
       " array([-5.31043823e+02,  3.63627090e+01,  2.88122444e+01, ...,\n",
       "         5.90820312e-02,  1.35567100e-02,  1.79849237e+01]),\n",
       " array([-4.86909637e+02,  8.13058540e-02, -5.44578910e-01, ...,\n",
       "         6.34765625e-02,  1.68571811e-02,  1.83794796e+01]),\n",
       " array([-5.18940613e+02,  5.14535446e+01,  2.96678581e+01, ...,\n",
       "         2.49023438e-02,  1.13213841e-02,  2.38687336e+01]),\n",
       " array([-6.27242615e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         9.37500000e-02,  2.15004766e-02,  1.24034613e+01]),\n",
       " array([-6.12557373e+02,  6.41938090e-01,  6.40358448e-01, ...,\n",
       "         3.36914062e-02,  2.92595397e-02,  1.66568264e+01]),\n",
       " array([-4.56737274e+02,  8.45813065e+01,  3.34042587e+01, ...,\n",
       "         6.25000000e-02,  1.24027079e-02,  7.92256892e+00]),\n",
       " array([-3.32850891e+02,  1.77575989e+02, -8.20902252e+01, ...,\n",
       "         9.27734375e-02,  1.38364214e-02,  2.07588941e+01]),\n",
       " array([-2.54043747e+02,  1.30134918e+02, -2.71058846e+01, ...,\n",
       "         6.98242188e-02,  2.64571859e-02,  1.29850388e+01]),\n",
       " array([-5.74510376e+02,  7.73657684e+01,  2.90992699e+01, ...,\n",
       "         1.01074219e-01,  3.43473716e-02,  3.94400120e+01]),\n",
       " array([-6.34203064e+02,  5.57467422e+01,  3.86585388e+01, ...,\n",
       "         2.88085938e-01,  1.60781249e-02,  1.63779303e+01]),\n",
       " array([-4.72837250e+02,  9.20081329e+01,  1.68731728e+01, ...,\n",
       "         2.27539062e-01,  6.98538321e-04,  6.04037225e+01]),\n",
       " array([-3.69007629e+02, -6.25225842e-01, -1.00436844e+02, ...,\n",
       "         1.11328125e-01,  4.82227286e-03,  1.45344615e+01]),\n",
       " array([-3.05076202e+02, -7.05688171e+01, -2.14786606e+01, ...,\n",
       "         8.30078125e-02,  1.08284994e-02,  1.61996901e+01]),\n",
       " array([-1.14231880e+02,  7.59800797e+01, -5.93419418e+01, ...,\n",
       "         5.76171875e-02,  9.13240892e-04,  4.16779369e+00]),\n",
       " array([-3.03433380e+02,  3.84268723e+01,  3.23661728e+01, ...,\n",
       "         5.95703125e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-5.21818726e+02,  3.39021759e+01,  3.55610943e+00, ...,\n",
       "         6.15234375e-02,  3.47703355e-02,  2.18866855e+01]),\n",
       " array([-3.36996674e+02,  1.51608601e+01, -1.18923912e+02, ...,\n",
       "         1.84082031e-01,  8.75581381e-03,  1.60251588e+01]),\n",
       " array([-83.75408173,  25.81352806,  -2.60709858, ...,   0.38964844,\n",
       "          0.        ,   0.        ]),\n",
       " array([-4.87669708e+02,  4.98834991e+01,  2.70885897e+00, ...,\n",
       "         3.22265625e-02,  1.52661802e-02,  2.36405626e+01]),\n",
       " array([-4.83154877e+02, -1.53844762e+00, -8.22232361e+01, ...,\n",
       "         1.45019531e-01,  3.80217009e-03,  1.70141265e+01]),\n",
       " array([-5.57652527e+02,  6.31613998e+01,  2.83956680e+01, ...,\n",
       "         9.37500000e-02,  2.13403554e-02,  2.13243634e+01]),\n",
       " array([-1.72894470e+02,  8.26571045e+01, -4.68746414e+01, ...,\n",
       "         4.29687500e-02,  1.71160852e-02,  3.31325948e+01]),\n",
       " array([-3.17829132e+02,  1.04576866e+02, -4.36138153e+01, ...,\n",
       "         9.27734375e-02,  1.17918762e-02,  1.90403223e+01]),\n",
       " array([-4.70225922e+02,  8.55347137e+01,  4.14850044e+01, ...,\n",
       "         5.46875000e-02,  6.90324860e-03,  2.09912673e+01]),\n",
       " array([-2.37250565e+02,  1.07412659e+02,  3.11151543e+01, ...,\n",
       "         1.35253906e-01,  1.28666279e-02,  1.75062835e+01]),\n",
       " array([-1.21294243e+02,  1.21646004e+02, -6.08967590e+01, ...,\n",
       "         6.59179688e-02,  1.43551800e-02,  1.82476640e+01]),\n",
       " array([-4.56971924e+02,  3.45012817e+01, -5.33678579e+00, ...,\n",
       "         2.25097656e-01,  1.37266246e-02,  2.67655224e+01]),\n",
       " array([-4.17076050e+02,  1.34997208e+02,  3.86892776e+01, ...,\n",
       "         1.36718750e-01,  3.92917280e-03,  1.35246828e+01]),\n",
       " array([-4.12609222e+02,  1.28346756e+02, -1.38635039e+00, ...,\n",
       "         6.83593750e-02,  9.56968330e-03,  2.45347142e+01]),\n",
       " array([-3.84068939e+02,  5.70902519e+01, -6.59542274e+00, ...,\n",
       "         1.00097656e-01,  1.00416851e-02,  2.61329025e+01]),\n",
       " array([-2.95050964e+02,  1.25548391e+01,  1.52857580e+01, ...,\n",
       "         3.71093750e-01,  2.88577194e-02,  2.10413784e+01]),\n",
       " array([-2.50212387e+02,  4.77330685e+00,  7.58863602e+01, ...,\n",
       "         1.66015625e-02,  1.69283260e-02,  1.87152758e+01]),\n",
       " array([-2.17628876e+02,  1.13909805e+02,  1.69869938e+01, ...,\n",
       "         1.37207031e-01,  9.43087323e-03,  2.76347220e+01]),\n",
       " array([-3.87645050e+02,  9.71273651e+01,  1.92713699e+01, ...,\n",
       "         2.58789062e-02,  1.79942762e-02,  1.26623616e+01]),\n",
       " array([-4.93313568e+02, -2.80169258e+01,  1.94109764e+01, ...,\n",
       "         4.93164062e-02,  1.89954574e-02,  1.97543651e+01]),\n",
       " array([-5.00557800e+02,  1.27664528e+02,  4.30027847e+01, ...,\n",
       "         4.19921875e-02,  1.49148097e-02,  2.02935517e+01]),\n",
       " array([-2.95375702e+02,  1.47824432e+02, -1.40158272e+01, ...,\n",
       "         1.03027344e-01,  2.40975320e-02,  2.65491754e+01]),\n",
       " array([-2.29062805e+02, -1.26887870e+00,  8.26670990e+01, ...,\n",
       "         9.86328125e-02,  4.13177457e-02,  2.70784885e+01]),\n",
       " array([-2.63018433e+02,  6.72963867e+01, -3.72019172e+00, ...,\n",
       "         4.77539062e-01,  2.69532815e-02,  2.07498536e+01]),\n",
       " array([-2.19906128e+02, -1.93706512e+01,  5.89577370e+01, ...,\n",
       "         4.68750000e-02,  1.75453363e-02,  2.11436376e+01]),\n",
       " array([-4.81958771e+02,  5.44159088e+01,  3.95757751e+01, ...,\n",
       "         1.46484375e-02,  1.91297138e-02,  3.35743427e+01]),\n",
       " array([-2.35643631e+02,  1.43704376e+02, -8.67563438e+00, ...,\n",
       "         1.44531250e-01,  2.17784548e-02,  2.10744649e+01]),\n",
       " array([-4.06891479e+02,  1.35837555e+02,  1.67059116e+01, ...,\n",
       "         1.72851562e-01,  1.35149850e-02,  1.16060816e+01]),\n",
       " array([-2.34967163e+02,  2.70969315e+01, -1.69811821e+01, ...,\n",
       "         1.79687500e-01,  6.83792685e-03,  2.59111792e+01]),\n",
       " array([-1.36597519e+02,  6.79409637e+01,  9.04493332e+00, ...,\n",
       "         1.50390625e-01,  1.74795370e-02,  2.10777268e+01]),\n",
       " array([-1.98912659e+02,  1.44022614e+02, -8.67271042e+00, ...,\n",
       "         1.56250000e-01,  1.31010149e-02,  2.12947994e+01]),\n",
       " array([-3.00848145e+02, -2.85069046e+01,  5.41222000e+01, ...,\n",
       "         3.74511719e-01,  2.52413928e-02,  1.49566337e+01]),\n",
       " array([-2.81289520e+02,  9.37862396e+01, -1.67037315e+01, ...,\n",
       "         9.61914062e-02,  1.23667249e-02,  1.71645582e+01]),\n",
       " array([-3.00998657e+02,  1.16178112e+01, -7.57347260e+01, ...,\n",
       "         6.25000000e-02,  2.54794794e-02,  2.20173940e+01]),\n",
       " array([-4.45259491e+02,  1.05727402e+02,  1.30042391e+01, ...,\n",
       "         7.76367188e-02,  1.79978060e-02,  2.53295600e+01]),\n",
       " array([-5.69593201e+02,  6.83384857e+01,  8.76914501e+00, ...,\n",
       "         9.03320312e-02,  1.54034700e-02,  1.46362543e+01]),\n",
       " array([-5.07522949e+02,  5.54603157e+01,  2.10984592e+01, ...,\n",
       "         9.96093750e-02,  1.13958355e-02,  1.94920689e+01]),\n",
       " array([-4.87548920e+02,  5.92454147e+00, -6.76454639e+00, ...,\n",
       "         7.37304688e-02,  6.82530249e-03,  1.51150852e+01]),\n",
       " array([-1.63734680e+02,  1.59481079e+02, -3.59738274e+01, ...,\n",
       "         1.62597656e-01,  2.54161474e-02,  1.69608682e+01]),\n",
       " array([-1.55658875e+02,  1.20420059e+02,  1.36089983e+01, ...,\n",
       "         2.08496094e-01,  1.06289054e-02,  1.10078983e+01]),\n",
       " array([-4.62109497e+02,  7.84076004e+01,  5.38236732e+01, ...,\n",
       "         4.68750000e-02,  1.35037927e-02,  2.34266043e+01]),\n",
       " array([-1.45872513e+02,  2.65926781e+01,  4.79223900e+01, ...,\n",
       "         1.67968750e-01,  1.46096167e-02,  2.82662779e+01]),\n",
       " array([-3.88197357e+02,  1.28919250e+02,  6.06856406e-01, ...,\n",
       "         9.03320312e-02,  1.17485368e-02,  1.36300504e+01]),\n",
       " array([-4.60901459e+02,  7.07858734e+01,  4.10736351e+01, ...,\n",
       "         8.39843750e-02,  1.26381927e-02,  1.93477020e+01]),\n",
       " array([-4.02952118e+02,  6.21649017e+01,  1.95937767e+01, ...,\n",
       "         1.41601562e-02,  5.80761447e-03,  6.78424910e+00]),\n",
       " array([-6.12062134e+02,  2.18587732e+00,  2.01107550e+00, ...,\n",
       "         3.56445312e-02,  3.07724500e-02,  2.34354466e+01]),\n",
       " array([-4.61349091e+02,  1.14105118e+02,  1.47719574e+01, ...,\n",
       "         1.71386719e-01,  1.43195760e-02,  2.35916555e+01]),\n",
       " array([-1.96250610e+02,  1.43799530e+02, -7.58100281e+01, ...,\n",
       "         3.85742188e-02,  8.59493631e-03,  1.72309175e+01]),\n",
       " array([-2.27191711e+02,  9.27181244e+01,  6.29147100e+00, ...,\n",
       "         2.29492188e-02,  2.75158100e-02,  1.92181394e+01]),\n",
       " array([-5.32283142e+02,  5.17721901e+01,  3.13473015e+01, ...,\n",
       "         3.85742188e-02,  3.98564726e-02,  2.23445311e+01]),\n",
       " array([-3.00121338e+02,  1.24116196e+02, -6.21503906e+01, ...,\n",
       "         2.46582031e-01,  2.02111349e-02,  7.07001090e+01]),\n",
       " array([-2.35467285e+02,  1.12632149e+02, -2.90647507e+01, ...,\n",
       "         1.66992188e-01,  4.54755948e-02,  3.20330888e+01]),\n",
       " array([-2.41929321e+02,  1.51352036e+02,  3.06236801e+01, ...,\n",
       "         7.32421875e-02,  1.42209281e-02,  2.63162136e+01]),\n",
       " array([-1.33021423e+02,  1.18116524e+02,  2.64488106e+01, ...,\n",
       "         2.92968750e-02,  1.33504222e-02,  7.56855309e+00]),\n",
       " array([-1.22048798e+02,  1.37339874e+02, -2.87339249e+01, ...,\n",
       "         9.27734375e-02,  9.74827346e-03,  2.04031557e+01]),\n",
       " array([-5.12850281e+02,  5.43021393e+01,  4.15923691e+01, ...,\n",
       "         6.88476562e-02,  4.08441618e-02,  1.85308278e+01]),\n",
       " array([-1.87136017e+02,  1.20193253e+02, -3.91389704e+00, ...,\n",
       "         1.04492188e-01,  1.15019383e-03,  2.92692691e+01]),\n",
       " array([-5.23111816e+02,  2.94687080e+01,  2.53474693e+01, ...,\n",
       "         1.13769531e-01,  1.65759714e-02,  2.15846628e+01]),\n",
       " array([-1.93770523e+02,  1.47391113e+02, -8.24565887e+00, ...,\n",
       "         1.64550781e-01,  1.53455675e-02,  1.50617748e+01]),\n",
       " array([-2.07746887e+02,  8.78064651e+01, -5.36153908e+01, ...,\n",
       "         5.81054688e-02,  1.03655298e-02,  2.43363321e+01]),\n",
       " array([-1.75082764e+02,  1.49829071e+02, -2.78119240e+01, ...,\n",
       "         1.95312500e-02,  1.28764812e-02,  1.68023646e+01]),\n",
       " array([-4.77491028e+02, -3.87381058e+01, -1.72770958e+01, ...,\n",
       "         1.19628906e-01,  6.89368871e-03,  1.71484709e+01]),\n",
       " array([-1.70569122e+02,  1.17660370e+02, -6.94091187e+01, ...,\n",
       "         1.50390625e-01,  9.42623326e-03,  1.68304220e+01]),\n",
       " array([-2.39469940e+02,  7.79287338e+01, -1.40201941e+01, ...,\n",
       "         5.66406250e-02,  1.29552381e-02,  2.04264998e+01]),\n",
       " array([-1.70734970e+02,  9.52683258e+01,  1.33362389e+01, ...,\n",
       "         2.00195312e-02,  8.64670000e-03,  1.72927767e+01]),\n",
       " array([-3.90995178e+02, -1.42305145e+01, -2.09496593e+01, ...,\n",
       "         4.57031250e-01,  2.79597386e-02,  2.15629041e+01]),\n",
       " array([-1.55950638e+02, -7.99633560e+01,  2.66621227e+01, ...,\n",
       "         6.29882812e-02,  1.11556837e-02,  1.64311111e+01]),\n",
       " array([-2.17784134e+02,  1.28775742e+02,  3.48982506e+01, ...,\n",
       "         1.22558594e-01,  1.27125058e-02,  1.97362095e+01]),\n",
       " array([-5.18022949e+02,  1.04850616e+02,  3.86920776e+01, ...,\n",
       "         1.26953125e-01,  1.39948660e-02,  2.04822779e+01]),\n",
       " array([-5.65293701e+02, -1.64061880e+00,  7.57164192e+00, ...,\n",
       "         8.93554688e-02,  8.67278701e-03,  1.77128673e+01]),\n",
       " array([-5.56693970e+02,  3.73183975e+01,  3.26169891e+01, ...,\n",
       "         7.08007812e-02,  9.63008849e-03,  2.00239792e+01]),\n",
       " array([-6.94363220e+02,  3.76906891e+01,  1.62620983e+01, ...,\n",
       "         6.20117188e-02,  8.36059934e-03,  1.75446689e+01]),\n",
       " array([-3.39793427e+02,  7.13190536e+01, -5.37895432e+01, ...,\n",
       "         2.32910156e-01,  4.14410690e-03,  2.28702322e+01]),\n",
       " array([-3.90277527e+02,  3.30820503e+01,  1.43654346e+01, ...,\n",
       "         3.06152344e-01,  2.90292672e-02,  1.82181701e+01]),\n",
       " array([-2.39447281e+02,  8.22240448e+01, -1.23651752e+01, ...,\n",
       "         6.54296875e-02,  1.84306503e-02,  2.15253562e+01]),\n",
       " array([-5.69571289e+02,  4.67993431e+01,  2.48064709e+01, ...,\n",
       "         2.09960938e-01,  8.15059384e-03,  2.75103658e+01]),\n",
       " array([-2.54650467e+02,  1.62260773e+02, -1.68782005e+01, ...,\n",
       "         1.16210938e-01,  9.59153855e-03,  1.72387227e+01]),\n",
       " array([-5.30257141e+02,  2.04076691e+01, -9.41240692e+00, ...,\n",
       "         8.00781250e-02,  4.85216355e-03,  1.06330872e+01]),\n",
       " array([-4.05547211e+02,  1.18779724e+02,  4.06271210e+01, ...,\n",
       "         4.39453125e-02,  1.35852637e-03,  7.68206418e+00]),\n",
       " array([-3.82232727e+02,  7.45551300e+01, -1.55505295e+01, ...,\n",
       "         1.69433594e-01,  1.90896024e-02,  2.46919736e+01]),\n",
       " array([-4.64746307e+02, -1.65557232e+01,  5.47328033e+01, ...,\n",
       "         5.02929688e-02,  6.05021264e-03,  1.69281930e+01]),\n",
       " array([-1.96505600e+02,  1.40035645e+02, -5.55408859e+01, ...,\n",
       "         5.85937500e-02,  1.54104416e-02,  2.57103533e+01]),\n",
       " array([-5.07921417e+02,  4.94881809e-01,  4.94434536e-01, ...,\n",
       "         8.64257812e-02,  2.58854709e-03,  1.27915800e+01]),\n",
       " array([-5.65708374e+02,  3.19120522e+01,  1.43990517e+01, ...,\n",
       "         5.37109375e-02,  1.50531821e-02,  2.04523876e+01]),\n",
       " array([ 4.38593102e+01,  4.69187737e+01, -3.13580704e+01, ...,\n",
       "         9.61914062e-02,  2.18174866e-02,  2.24204630e+01]),\n",
       " array([-1.38697479e+02,  8.94412155e+01, -8.77981796e+01, ...,\n",
       "         9.52148438e-02,  1.36366775e-02,  2.08273917e+01]),\n",
       " array([-1.12430428e+02,  1.02653198e+02, -5.76779556e+01, ...,\n",
       "         1.93359375e-01,  2.93501977e-02,  2.47906685e+01]),\n",
       " array([-2.57630188e+02,  1.40766815e+02, -3.17954564e+00, ...,\n",
       "         7.42187500e-02,  1.86808140e-02,  2.28090122e+01]),\n",
       " array([-3.11516693e+02,  8.88055420e+01, -3.16033821e+01, ...,\n",
       "         2.05078125e-02,  2.42465402e-02,  2.24231988e+01]),\n",
       " array([-3.93441772e+02, -7.97445297e-01, -4.66135254e+01, ...,\n",
       "         7.61718750e-02,  9.11271344e-03,  7.39087835e+00]),\n",
       " array([-4.73456390e+02,  2.64089336e+01, -1.93690147e+01, ...,\n",
       "         1.79199219e-01,  7.50647710e-03,  1.49879709e+01]),\n",
       " array([-1.40847137e+02,  3.52498550e+01, -2.97656155e+01, ...,\n",
       "         9.57031250e-02,  1.68921905e-02,  2.17012197e+01]),\n",
       " array([-3.16631012e+02,  1.51178696e+02, -1.74060211e+01, ...,\n",
       "         2.83203125e-02,  1.59581285e-02,  1.45570993e+01]),\n",
       " array([-3.68643829e+02,  6.27888718e+01,  4.06470947e+01, ...,\n",
       "         4.58984375e-02,  1.51236492e-02,  2.01925635e+01]),\n",
       " array([-5.92599792e+02,  4.85251007e+01,  2.02177639e+01, ...,\n",
       "         5.66406250e-02,  1.49951869e-02,  2.17672199e+01]),\n",
       " array([-6.02409424e+02,  4.35487652e+00,  3.44967985e+00, ...,\n",
       "         5.81054688e-02,  8.98265142e-03,  1.86041862e+01]),\n",
       " array([-4.70059845e+02,  1.84394016e+01, -7.81145859e+00, ...,\n",
       "         5.41992188e-02,  8.35897146e-03,  2.04760298e+01]),\n",
       " array([-106.04724121,   53.78616333,  -33.36463165, ...,    0.38623047,\n",
       "           0.        ,    0.        ]),\n",
       " array([-5.72088135e+02,  9.37589836e+00,  7.03981209e+00, ...,\n",
       "         8.59375000e-02,  1.22549359e-02,  2.21413657e+01]),\n",
       " array([-5.33976868e+02,  4.91271305e+00,  3.13069701e+00, ...,\n",
       "         1.59179688e-01,  8.50659974e-03,  1.62713796e+01]),\n",
       " array([-5.93513367e+02,  2.01058979e+01,  1.45880728e+01, ...,\n",
       "         5.85937500e-02,  2.95888034e-02,  1.40888572e+01]),\n",
       " array([-4.61489319e+02,  3.88340149e+01, -3.09162521e+01, ...,\n",
       "         6.05468750e-02,  8.39676541e-03,  1.50235534e+01]),\n",
       " array([-3.84454620e+02,  1.20215553e+02, -2.59447308e+01, ...,\n",
       "         5.17578125e-02,  1.34701506e-02,  1.48126766e+01]),\n",
       " array([-2.62000732e+02,  7.93256836e+01, -7.50699081e+01, ...,\n",
       "         1.45996094e-01,  3.43541910e-02,  1.69422552e+01]),\n",
       " array([-1.94848404e+02,  1.08541000e+02,  2.31179447e+01, ...,\n",
       "         1.53808594e-01,  9.31207022e-03,  1.21436931e+01]),\n",
       " array([-1.98373978e+02,  5.52786446e+01, -8.18807316e+00, ...,\n",
       "         7.03125000e-02,  1.73176943e-02,  1.98033243e+01]),\n",
       " array([-4.24849548e+02,  1.10061295e+02,  2.00683556e+01, ...,\n",
       "         1.12304688e-02,  2.97003020e-02,  1.76117435e+01]),\n",
       " array([-5.23142822e+02,  2.51669884e+01, -3.38840652e+00, ...,\n",
       "         2.61230469e-01,  8.31082648e-03,  1.30703509e+01]),\n",
       " array([-5.15836792e+02,  8.14087391e+00, -9.94965649e+00, ...,\n",
       "         1.65039062e-01,  6.07634336e-03,  1.65471375e+01]),\n",
       " array([-1.46512512e+02,  1.06395767e+02, -9.63027668e+00, ...,\n",
       "         1.66015625e-02,  1.37036875e-02,  1.75314084e+01]),\n",
       " array([-5.81274048e+02,  4.85092545e+01,  1.93276215e+01, ...,\n",
       "         2.01660156e-01,  1.43635231e-02,  3.33232343e+01]),\n",
       " array([-5.07071564e+02,  5.35977554e+00, -9.48095989e+00, ...,\n",
       "         4.58984375e-02,  8.51279560e-03,  1.41444430e+01]),\n",
       " array([-5.01979828e+02,  3.46754456e+00,  3.06102943e+00, ...,\n",
       "         3.95507812e-02,  2.02202299e-02,  2.70259827e+01]),\n",
       " array([-6.21116760e+02, -6.45783567e+00, -1.10864964e+01, ...,\n",
       "         2.11425781e-01,  9.67409930e-03,  2.40993649e+01]),\n",
       " array([-5.02346619e+02,  3.41174698e+01,  2.50688744e+01, ...,\n",
       "         1.50390625e-01,  1.02636491e-02,  2.92479008e+01]),\n",
       " array([-1.21499161e+02,  8.01168365e+01, -4.27195740e+01, ...,\n",
       "         1.89453125e-01,  7.15134531e-03,  1.83191344e+01]),\n",
       " array([-5.86321594e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         8.20312500e-02,  3.13520678e-02,  1.89693645e+01]),\n",
       " array([-4.49100311e+02,  1.15960922e+02,  3.76880760e+01, ...,\n",
       "         2.39257812e-02,  2.33546437e-02,  1.65579349e+01]),\n",
       " array([-5.48121277e+02,  4.82931213e+01,  3.35426483e+01, ...,\n",
       "         9.08203125e-02,  3.29277160e-02,  2.84329265e+01]),\n",
       " array([-3.89624084e+02,  1.00202103e+02,  1.90790138e+01, ...,\n",
       "         1.37695312e-01,  9.60019562e-03,  2.61867106e+01]),\n",
       " array([-3.29912689e+02,  1.29668640e+02,  1.76443024e+01, ...,\n",
       "         6.39648438e-02,  1.87616537e-02,  2.14114562e+01]),\n",
       " array([-1.59904205e+02,  5.76494141e+01, -2.73781548e+01, ...,\n",
       "         1.06445312e-01,  6.54204971e-03,  1.84779257e+01]),\n",
       " array([-3.39256866e+02, -1.87648451e+00,  1.04304600e+01, ...,\n",
       "         1.68945312e-01,  7.12838196e-03,  2.65635610e+01]),\n",
       " array([-5.54039490e+02,  7.68790131e+01,  2.54662628e+01, ...,\n",
       "         2.19726562e-01,  7.14482720e-03,  1.27304539e+01]),\n",
       " array([-3.37282562e+02,  1.08447906e+02,  2.89257584e+01, ...,\n",
       "         9.57031250e-02,  1.08179667e-02,  2.68259525e+01]),\n",
       " array([-4.97828156e+02,  7.55355225e+01,  3.52399979e+01, ...,\n",
       "         1.61132812e-02,  1.67924615e-02,  2.30831161e+01]),\n",
       " array([-1.28136673e+02, -4.17378159e+01, -3.65346794e+01, ...,\n",
       "         1.90917969e-01,  4.48869541e-03,  3.05239737e+01]),\n",
       " array([-3.14541016e+02,  1.31005371e+02,  1.62306290e+01, ...,\n",
       "         3.85742188e-02,  2.67032170e-02,  1.37847483e+01]),\n",
       " array([-9.56620026e+01,  1.20578308e+02,  1.33725986e+01, ...,\n",
       "         1.10351562e-01,  1.38346834e-02,  1.20471627e+01]),\n",
       " array([-2.48546783e+02, -3.64004974e+01,  2.72262764e+01, ...,\n",
       "         3.07617188e-02,  3.25797137e-03,  7.80981332e+00]),\n",
       " array([-3.92990112e+02,  5.31856041e+01, -4.80879288e+01, ...,\n",
       "         1.89941406e-01,  1.12010526e-02,  2.15103596e+01]),\n",
       " array([-3.09241547e+02,  9.88227844e+01,  2.73378611e+00, ...,\n",
       "         6.73828125e-02,  1.53815235e-02,  1.00579783e+01]),\n",
       " array([-1.58223251e+02,  9.48101654e+01, -2.78247929e+00, ...,\n",
       "         6.39648438e-02,  2.66493236e-02,  2.75399745e+01]),\n",
       " array([-1.22942101e+02, -1.30651459e+02,  1.65134239e+01, ...,\n",
       "         1.43066406e-01,  5.15286823e-03,  1.77863538e+01]),\n",
       " array([-2.46003815e+02, -4.43880692e+01, -1.47712393e+01, ...,\n",
       "         1.82128906e-01,  1.38629630e-02,  2.13459566e+01]),\n",
       " array([-3.67256073e+02,  1.48505814e+02,  2.04204273e+01, ...,\n",
       "         4.29687500e-02,  1.20858822e-02,  1.25820622e+01]),\n",
       " array([-1.76937210e+02,  1.06618614e+02, -5.73806000e+01, ...,\n",
       "         3.44726562e-01,  1.26458106e-02,  1.70435503e+01]),\n",
       " array([-3.27530762e+02,  3.47398300e+01, -4.10496216e+01, ...,\n",
       "         1.64062500e-01,  1.28137647e-02,  1.40438423e+01]),\n",
       " array([-5.89680969e+02,  3.36083755e+01,  2.66327896e+01, ...,\n",
       "         1.71875000e-01,  3.83464388e-03,  1.35708436e+01]),\n",
       " array([-4.49263123e+02,  9.72845154e+01,  5.77152786e+01, ...,\n",
       "         2.09960938e-02,  4.63858494e-02,  1.93479404e+01]),\n",
       " array([-5.87925293e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         5.12695312e-02,  2.37908614e-02,  1.79025501e+01]),\n",
       " array([-4.99834717e+02,  4.53115692e+01, -1.44479847e+01, ...,\n",
       "         9.71679688e-02,  2.21835846e-02,  1.61087275e+01]),\n",
       " array([-2.12860672e+02,  1.37074677e+02, -7.11900101e+01, ...,\n",
       "         9.96093750e-02,  1.72153560e-02,  3.19782645e+01]),\n",
       " array([-6.29405518e+02,  3.24925385e+01,  4.34591961e+00, ...,\n",
       "         7.51953125e-02,  1.72417355e-02,  2.33255133e+01]),\n",
       " array([-3.27881866e+02,  1.17430786e+02,  2.69962559e+01, ...,\n",
       "         2.47070312e-01,  1.75644488e-02,  2.48484582e+01]),\n",
       " array([-6.77851013e+02,  1.93261223e+01,  1.54924641e+01, ...,\n",
       "         1.11328125e-01,  1.69922870e-02,  2.54743516e+01]),\n",
       " array([-5.95583008e+02,  5.12224722e+00,  2.87242079e+00, ...,\n",
       "         9.86328125e-02,  8.48761667e-03,  7.78908655e+00]),\n",
       " array([-2.32364120e+02,  9.33112488e+01, -2.68720436e+01, ...,\n",
       "         1.28417969e-01,  9.53925612e-03,  1.94126755e+01]),\n",
       " array([-6.53177368e+02,  4.92237244e+01, -6.09190063e+01, ...,\n",
       "         4.58984375e-02,  1.65103670e-02,  1.10259019e+01]),\n",
       " array([-2.82904022e+02,  2.18197823e+01,  5.94936037e+00, ...,\n",
       "         1.47460938e-01,  2.53250867e-02,  1.80927113e+01]),\n",
       " array([-1.95997833e+02,  1.11803589e+02,  2.94230595e+01, ...,\n",
       "         1.00585938e-01,  9.97673338e-03,  2.13725001e+01]),\n",
       " array([-5.77713806e+02,  5.69910049e+01,  2.32100544e+01, ...,\n",
       "         1.80664062e-02,  6.59176139e-03,  1.44000128e+01]),\n",
       " array([-4.06071869e+02,  1.06829453e+02,  1.35377765e+00, ...,\n",
       "         3.85742188e-02,  9.51933983e-03,  1.55143812e+01]),\n",
       " array([-5.31626282e+02,  2.75159225e+01,  1.89323463e+01, ...,\n",
       "         1.47949219e-01,  4.07476696e-03,  1.17890142e+01]),\n",
       " array([-2.23619415e+02,  1.30899338e+02, -2.32007637e+01, ...,\n",
       "         6.39648438e-02,  1.95637705e-02,  2.25107923e+01]),\n",
       " array([-1.22623833e+02,  1.21605537e+02, -9.47189026e+01, ...,\n",
       "         1.35253906e-01,  6.87861930e-03,  1.99441269e+01]),\n",
       " array([-4.98433929e+02,  5.36206589e+01,  1.97447453e+01, ...,\n",
       "         4.44335938e-02,  2.71989420e-02,  1.73267543e+01]),\n",
       " array([-2.45854904e+02,  4.05655289e+01, -4.78848343e+01, ...,\n",
       "         1.46484375e-01,  5.51328693e-03,  1.96494028e+01]),\n",
       " array([-1.77216019e+02,  1.57302124e+02, -1.88489151e+01, ...,\n",
       "         3.51562500e-02,  1.34025746e-02,  1.64923400e+01]),\n",
       " array([-4.32467590e+02,  6.44023895e+01,  1.21442866e+00, ...,\n",
       "         6.49414062e-02,  6.97904872e-03,  1.81169346e+01]),\n",
       " array([-4.31768707e+02,  1.06699173e+02,  4.05057526e+01, ...,\n",
       "         2.50976562e-01,  1.26513518e-02,  1.78137407e+01]),\n",
       " array([-2.51906570e+02,  1.06804626e+02, -1.81023655e+01, ...,\n",
       "         6.54296875e-02,  2.24269733e-02,  1.82587489e+01]),\n",
       " array([-5.54536560e+02,  2.42659760e+01,  2.32856369e+01, ...,\n",
       "         7.47070312e-02,  1.15279262e-02,  1.25013188e+01]),\n",
       " array([-3.61452454e+02,  1.52421551e+01, -6.11426468e+01, ...,\n",
       "         1.15722656e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-4.57928589e+02,  1.03184898e+02,  3.33637772e+01, ...,\n",
       "         8.00781250e-02,  7.20396153e-03,  4.20652121e+01]),\n",
       " array([-5.88675537e+02,  7.57450790e+01,  1.59365492e+01, ...,\n",
       "         8.83789062e-02,  2.15181451e-03,  1.29890010e+01]),\n",
       " array([-6.21761597e+02,  4.81603546e+01,  2.78946514e+01, ...,\n",
       "         1.15722656e-01,  4.58593276e-02,  2.03214049e+01]),\n",
       " array([-5.88289795e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         4.49218750e-02,  1.81199894e-02,  1.21160336e+01]),\n",
       " array([-2.62736816e+02,  8.40383911e+01, -1.56323967e+01, ...,\n",
       "         1.32812500e-01,  2.44850564e-02,  1.24468125e+01]),\n",
       " array([-4.78439819e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         4.73632812e-02,  5.76926171e-03,  2.05843329e+01]),\n",
       " array([-2.06398483e+02,  8.81752319e+01,  4.15916977e+01, ...,\n",
       "         5.17578125e-02,  1.46141667e-02,  3.10148865e+01]),\n",
       " array([-1.44568832e+02,  1.63419403e+02, -3.48241730e+01, ...,\n",
       "         6.00585938e-02,  3.24036320e-02,  2.27629587e+01]),\n",
       " array([-2.52841461e+02,  9.12952347e+01,  1.99678459e+01, ...,\n",
       "         4.29687500e-02,  1.32011482e-02,  1.85902745e+01]),\n",
       " array([-5.25707886e+02,  2.24423103e+01,  6.99587536e+00, ...,\n",
       "         2.00195312e-01,  7.95426061e-03,  1.48617297e+01]),\n",
       " array([-3.78115021e+02,  1.17334351e+02, -1.23244467e+01, ...,\n",
       "         6.98242188e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.19428452e+02,  8.10030212e+01, -2.10247993e+01, ...,\n",
       "         6.29882812e-02,  1.09550071e-02,  2.86743045e+01]),\n",
       " array([-3.61672546e+02,  2.75639210e+01, -2.62815905e+00, ...,\n",
       "         1.56250000e-01,  1.71797063e-02,  1.62332907e+01]),\n",
       " array([-5.30635010e+02, -4.43523073e+00,  1.20559139e+01, ...,\n",
       "         3.29101562e-01,  5.21169109e-03,  1.91293985e+01]),\n",
       " array([-2.96229523e+02, -5.33406525e+01,  2.75318871e+01, ...,\n",
       "         3.90625000e-01,  1.49511030e-02,  2.41365835e+01]),\n",
       " array([-55.49352646,  23.19789314, -60.51129913, ...,   0.12451172,\n",
       "          0.        ,   0.        ]),\n",
       " array([-7.21752167e+01,  8.91748199e+01, -2.22113838e+01, ...,\n",
       "         3.12500000e-02,  8.36679792e-03,  1.62777007e+01]),\n",
       " array([-1.73214752e+02,  1.17745255e+02,  2.03509178e+01, ...,\n",
       "         6.59179688e-02,  2.13621409e-02,  1.46518916e+01]),\n",
       " array([-2.98990448e+02,  1.34047882e+02, -4.57466583e+01, ...,\n",
       "         1.13769531e-01,  1.25313370e-02,  2.27092654e+01]),\n",
       " array([-1.16462761e+02,  1.39309814e+02,  2.61756210e+01, ...,\n",
       "         7.27539062e-02,  1.50548626e-02,  1.66856006e+01]),\n",
       " array([-3.90551514e+02, -1.29128555e+02,  1.45395613e+01, ...,\n",
       "         2.25585938e-01,  5.25306587e-03,  1.43132225e+01]),\n",
       " array([-101.53274536,  -30.0200386 ,  -55.18317413, ...,    0.21972656,\n",
       "           0.        ,    0.        ]),\n",
       " array([-5.27975464e+02,  6.50462270e-01,  6.50315821e-01, ...,\n",
       "         2.39257812e-02,  1.02535012e-02,  2.71088749e+01]),\n",
       " array([-5.13756470e+02,  6.05253754e+01,  1.88490791e+01, ...,\n",
       "         1.16210938e-01,  2.67788299e-02,  2.29649216e+01]),\n",
       " array([-2.59482208e+02,  1.70482819e+02, -5.03518829e+01, ...,\n",
       "         1.73828125e-01,  1.02782617e-02,  2.56990343e+01]),\n",
       " array([-4.45406647e+02,  1.27164795e+02,  1.34092693e+01, ...,\n",
       "         1.26464844e-01,  2.33791355e-02,  1.27339259e+01]),\n",
       " array([-6.08239380e+02,  1.04058571e+01,  2.02355528e+00, ...,\n",
       "         7.71484375e-02,  1.71897832e-02,  3.10811847e+01]),\n",
       " array([-2.71619293e+02,  1.19844925e+02, -2.58484879e+01, ...,\n",
       "         1.69921875e-01,  1.18973846e-02,  3.40247393e+01]),\n",
       " array([-5.31754761e+02,  7.56000061e+01,  1.72750320e+01, ...,\n",
       "         1.04492188e-01,  1.56502762e-02,  2.08689883e+01]),\n",
       " array([-4.08798248e+02, -4.30420074e+01,  3.19014740e+01, ...,\n",
       "         3.36914062e-02,  4.35341106e-02,  2.23369092e+01]),\n",
       " array([-5.20478210e+02,  1.12016907e+02,  2.26598701e+01, ...,\n",
       "         1.91406250e-01,  7.40982250e-03,  1.79208398e+01]),\n",
       " array([-3.56843109e+02,  9.82238541e+01,  3.43995667e+01, ...,\n",
       "         1.71386719e-01,  9.87031001e-03,  1.66933730e+01]),\n",
       " array([-2.20345245e+02,  1.14301147e+02,  2.00142670e+01, ...,\n",
       "         7.37304688e-02,  5.80615517e-02,  2.35318482e+01]),\n",
       " array([-6.07298096e+02,  2.69620667e+01, -7.98261261e+00, ...,\n",
       "         1.67480469e-01,  1.46346496e-02,  1.80280298e+01]),\n",
       " array([-5.08338501e+02,  7.41003265e+01,  2.81739635e+01, ...,\n",
       "         1.22558594e-01,  5.68011762e-03,  1.64878890e+01]),\n",
       " array([-2.86689301e+02,  7.03277512e+01,  3.31435623e+01, ...,\n",
       "         5.66406250e-02,  1.20416253e-02,  1.72528476e+01]),\n",
       " array([-5.98204712e+02,  5.00265656e+01,  1.88005066e+01, ...,\n",
       "         3.51562500e-02,  2.64851850e-02,  2.55559534e+01]),\n",
       " array([-5.41678711e+02,  5.05782051e+01,  1.66335583e+01, ...,\n",
       "         8.44726562e-02,  2.67897992e-02,  3.84469807e+01]),\n",
       " array([-5.18586365e+02,  6.63008881e+01,  1.49555759e+01, ...,\n",
       "         1.24023438e-01,  1.77000519e-02,  2.24107385e+01]),\n",
       " array([-4.26515625e+02,  7.96386261e+01,  5.92544861e+01, ...,\n",
       "         4.34570312e-02,  2.82024749e-02,  2.62511194e+01]),\n",
       " array([-4.37036560e+02,  1.13406181e+02,  2.53874054e+01, ...,\n",
       "         1.46484375e-02,  4.74676128e-02,  1.74064755e+01]),\n",
       " array([-4.29974091e+02, -3.88889771e+01, -1.16271057e+01, ...,\n",
       "         1.06933594e-01,  7.57926529e-03,  2.32192278e+01]),\n",
       " array([-1.12071976e+02,  1.08756729e+02,  2.47686462e+01, ...,\n",
       "         2.29492188e-02,  1.55935633e-02,  1.04084469e+01]),\n",
       " array([-5.61788147e+02,  1.70676785e+01,  1.47307491e+01, ...,\n",
       "         3.45703125e-01,  7.62690727e-03,  1.98863119e+01]),\n",
       " array([-4.39131592e+02, -4.42458572e+01,  7.34094925e+01, ...,\n",
       "         3.08105469e-01,  2.83872479e-03,  7.54341602e+01]),\n",
       " array([-2.28277100e+02, -7.53123016e+01, -1.52408276e+01, ...,\n",
       "         7.91015625e-02,  8.51051816e-03,  2.91838706e+01]),\n",
       " array([-5.72902588e+02,  8.46302986e+00,  7.93575287e+00, ...,\n",
       "         1.19628906e-01,  8.04862047e-03,  1.56849429e+01]),\n",
       " array([-1.72066284e+02, -3.04462509e+01,  3.84367764e-01, ...,\n",
       "         2.35351562e-01,  1.06962074e-02,  2.35931128e+01]),\n",
       " array([-2.11141281e+02,  1.78380798e+02, -4.88980255e+01, ...,\n",
       "         6.49414062e-02,  1.74225031e-02,  1.82146132e+01]),\n",
       " array([-3.56834503e+02, -6.65484314e+01,  2.26482773e+01, ...,\n",
       "         2.60742188e-01,  1.08074080e-02,  1.45705849e+01]),\n",
       " array([-4.79688934e+02,  1.08606323e+02,  1.66484127e+01, ...,\n",
       "         4.34570312e-02,  6.71728610e-03,  1.19740203e+01]),\n",
       " array([-6.00732483e+02,  5.71719837e+00,  3.41668034e+00, ...,\n",
       "         9.57031250e-02,  1.10373365e-02,  3.23920041e+01]),\n",
       " array([-6.31864685e+02,  3.66944313e+01,  3.03443928e+01, ...,\n",
       "         1.05468750e-01,  5.78739139e-02,  2.52238959e+01]),\n",
       " array([-3.45786591e+02,  1.46350784e+02,  2.85534477e+01, ...,\n",
       "         9.08203125e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-2.15421219e+02,  4.90917282e+01, -3.62618332e+01, ...,\n",
       "         1.62597656e-01,  9.82650764e-03,  2.45233074e+01]),\n",
       " array([-5.14775818e+02,  3.43333130e+01,  1.11069412e+01, ...,\n",
       "         1.33300781e-01,  1.33135299e-02,  1.74199641e+01]),\n",
       " array([-4.73394440e+02,  1.18797054e+01, -8.86625385e+00, ...,\n",
       "         1.56738281e-01,  7.24969310e-03,  1.55736089e+01]),\n",
       " array([-2.48358612e+02,  1.18886208e+02,  3.28449364e+01, ...,\n",
       "         5.17578125e-02,  2.47351798e-02,  1.50988147e+01]),\n",
       " array([-7.33559387e+02,  2.17695415e-01,  2.15285152e-01, ...,\n",
       "         1.58203125e-01,  3.03241305e-02,  1.28046736e+01]),\n",
       " array([-2.27476212e+02,  1.54756088e+02, -3.14344311e+00, ...,\n",
       "         5.12695312e-02,  1.09578272e-02,  1.70695558e+01]),\n",
       " array([-5.63684753e+02,  7.26759491e+01,  3.13357029e+01, ...,\n",
       "         2.63671875e-02,  1.09676852e-02,  1.37370586e+01]),\n",
       " array([-2.77202911e+02,  7.98783569e+01,  3.11534233e+01, ...,\n",
       "         7.81250000e-02,  4.24014558e-03,  1.42962039e+01]),\n",
       " array([-2.78174530e+02,  1.08685135e+02,  1.95335102e+01, ...,\n",
       "         9.32617188e-02,  9.93652661e-03,  2.64128149e+01]),\n",
       " array([-3.06569763e+02,  1.53169373e+02,  4.41879005e+01, ...,\n",
       "         3.61328125e-02,  1.99516698e-02,  1.34649977e+01]),\n",
       " array([-5.00091064e+02,  8.25249710e+01,  6.33815117e+01, ...,\n",
       "         1.12304688e-02,  1.01009912e-02,  2.05348149e+01]),\n",
       " array([-3.15690643e+02, -2.24691620e+01,  8.32216034e+01, ...,\n",
       "         1.75781250e-02,  3.61254353e-02,  2.31983259e+01]),\n",
       " array([-3.03290710e+02, -5.73964767e+01, -6.21222572e+01, ...,\n",
       "         1.29882812e-01,  7.61704767e-03,  2.44492263e+01]),\n",
       " array([-1.68561615e+02,  5.67016983e+01, -1.92471294e+01, ...,\n",
       "         1.44531250e-01,  8.59009956e-03,  2.15580165e+01]),\n",
       " array([-4.09459351e+02,  1.62212418e+02,  3.70680656e+01, ...,\n",
       "         4.10156250e-02,  1.16879511e-02,  1.70087278e+01]),\n",
       " array([-5.10179688e+02,  4.13257065e+01,  9.32575989e+00, ...,\n",
       "         1.46484375e-02,  1.78503108e-02,  1.76120311e+01]),\n",
       " array([-4.43663696e+02,  9.24007645e+01, -5.13928452e+01, ...,\n",
       "         8.00781250e-02,  8.43540095e-03,  2.31464058e+01]),\n",
       " array([-3.14293823e+02,  1.01078308e+02, -1.52115259e+01, ...,\n",
       "         4.05273438e-02,  1.94181353e-02,  2.13318929e+01]),\n",
       " array([-1.32691391e+02,  1.01438599e+02, -9.37120514e+01, ...,\n",
       "         6.34765625e-02,  2.08793913e-02,  2.31346712e+01]),\n",
       " array([-3.19558685e+02,  7.64456482e+01, -7.21724243e+01, ...,\n",
       "         1.18652344e-01,  1.84092687e-02,  2.28715107e+01]),\n",
       " array([-5.31903320e+02, -1.87087212e+01,  7.36404991e+00, ...,\n",
       "         4.73632812e-02,  6.99291606e-03,  2.01816753e+01]),\n",
       " array([-4.20583344e+02,  5.96041870e+01,  3.78089218e+01, ...,\n",
       "         9.37500000e-02,  8.89418755e-03,  1.52411744e+01]),\n",
       " array([-1.38451004e+02,  1.56572083e+02, -3.07594357e+01, ...,\n",
       "         1.71875000e-01,  1.53215260e-02,  1.34628460e+01]),\n",
       " array([-1.41400574e+02,  1.07763107e+02,  7.02996135e-01, ...,\n",
       "         4.58984375e-02,  2.44302103e-02,  8.98868516e+00]),\n",
       " array([-5.14848450e+02,  5.18695688e+00,  2.23856335e+01, ...,\n",
       "         1.15234375e-01,  3.19109900e-02,  1.84438944e+01]),\n",
       " array([-4.42319977e+02,  9.93013000e+01, -2.30082397e+01, ...,\n",
       "         4.34570312e-02,  1.80342468e-02,  1.61630750e+01]),\n",
       " array([-1.19069916e+02,  1.21863953e+02,  7.15727234e+00, ...,\n",
       "         2.39257812e-02,  1.79788493e-02,  2.59749144e+01]),\n",
       " array([-4.91362732e+02,  7.89117813e+01,  5.03771667e+01, ...,\n",
       "         8.30078125e-02,  2.30891998e-02,  1.78335831e+01]),\n",
       " array([-2.37471283e+02,  7.85110397e+01,  1.32289085e+01, ...,\n",
       "         2.49023438e-02,  1.59154332e-02,  2.08245143e+01]),\n",
       " array([-5.32283325e+02,  1.59863882e+01, -1.05384102e+01, ...,\n",
       "         9.52148438e-02,  4.13300705e-03,  2.29127660e+01]),\n",
       " array([-5.98234497e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         3.95507812e-02,  2.45357010e-02,  8.92664567e+00]),\n",
       " array([-4.84265198e+02,  9.27592468e+01,  2.45897102e+01, ...,\n",
       "         2.14355469e-01,  2.19635292e-02,  1.74407318e+01]),\n",
       " array([-2.90708069e+02,  8.47599258e+01,  5.28046799e+01, ...,\n",
       "         2.63671875e-02,  8.39864353e-03,  9.64183658e+00]),\n",
       " array([-3.16781586e+02,  1.34637863e+02,  7.29236984e+00, ...,\n",
       "         8.39843750e-02,  2.25530540e-02,  1.43959105e+01]),\n",
       " array([-8.57048264e+01,  1.08700951e+02, -2.58053493e+01, ...,\n",
       "         9.57031250e-02,  1.63630118e-02,  2.14776695e+01]),\n",
       " array([-2.12091675e+02,  8.31313934e+01, -4.55435028e+01, ...,\n",
       "         2.07519531e-01,  1.81885024e-02,  2.22386688e+01]),\n",
       " array([-3.80832855e+02, -3.71390419e+01,  7.28237457e+01, ...,\n",
       "         7.51953125e-02,  2.30147754e-02,  2.29649693e+01]),\n",
       " array([-1.48275955e+02,  1.17652359e+02,  3.51216278e+01, ...,\n",
       "         1.60644531e-01,  7.65446206e-03,  1.31050944e+01]),\n",
       " array([-3.81715515e+02,  5.46462154e+00,  3.42228851e+01, ...,\n",
       "         3.15917969e-01,  1.15087738e-02,  2.88079202e+01]),\n",
       " array([-4.93574158e+02,  2.53642044e+01, -2.37238884e+01, ...,\n",
       "         1.46972656e-01,  8.69434609e-03,  1.57758087e+01]),\n",
       " array([-1.61666153e+02,  1.09867989e+02,  1.92415771e+01, ...,\n",
       "         3.07617188e-02,  8.34975578e-03,  1.85638934e+01]),\n",
       " array([-3.33524048e+02,  1.32705780e+02,  2.70420914e+01, ...,\n",
       "         1.61132812e-02,  2.24431787e-02,  2.16828018e+01]),\n",
       " array([-4.85744110e+02,  7.39636688e+01,  3.23118248e+01, ...,\n",
       "         1.47460938e-01,  1.24137047e-02,  9.16598737e+00]),\n",
       " array([-1.85880386e+02, -5.54428749e+01,  1.01688938e+01, ...,\n",
       "         2.29492188e-02,  1.25276972e-02,  1.54696509e+01]),\n",
       " array([-1.31775391e+02,  1.21713722e+02,  3.34149055e+01, ...,\n",
       "         1.70898438e-02,  4.18809816e-02,  1.61746964e+01]),\n",
       " array([-1.38044235e+02,  1.22740585e+02,  4.97759247e+00, ...,\n",
       "         2.35351562e-01,  2.21282666e-02,  2.20481291e+01]),\n",
       " array([-4.54228516e+02,  7.45067749e+01,  4.23524952e+00, ...,\n",
       "         1.02050781e-01,  9.54598344e-03,  2.21789867e+01]),\n",
       " array([-3.24380219e+02, -3.57423477e+01,  9.54684601e+01, ...,\n",
       "         6.25000000e-02,  7.92756149e-03,  1.11949004e+01]),\n",
       " array([-3.17063080e+02,  1.01519234e+02, -6.59320068e+01, ...,\n",
       "         6.20117188e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-4.52372406e+02,  1.30033325e+02,  1.31794014e+01, ...,\n",
       "         6.93359375e-02,  1.41525063e-02,  1.42721415e+01]),\n",
       " array([-2.38490677e+02,  7.19269257e+01, -9.77331829e+00, ...,\n",
       "         1.17675781e-01,  1.09328708e-02,  9.64551494e+00]),\n",
       " array([-3.34934296e+02,  6.41554947e+01,  1.83961439e+00, ...,\n",
       "         1.43554688e-01,  1.23904795e-02,  1.88718617e+01]),\n",
       " array([-2.82990601e+02,  6.30536423e+01, -4.27936783e+01, ...,\n",
       "         2.63671875e-02,  9.31127161e-03,  2.10428059e+01]),\n",
       " array([-3.36620941e+02,  1.50630341e+02,  4.88948584e+00, ...,\n",
       "         7.91015625e-02,  1.40799272e-02,  1.88369855e+01]),\n",
       " array([-3.74204956e+02,  1.77552032e+02, -3.49395294e+01, ...,\n",
       "         5.32226562e-02,  5.80774947e-03,  4.94195744e+00]),\n",
       " array([-5.22144409e+02,  8.72035065e+01,  4.95199127e+01, ...,\n",
       "         1.34277344e-01,  1.13087798e-02,  2.36891925e+01]),\n",
       " array([-4.39012787e+02,  6.50040665e+01, -1.24172506e+01, ...,\n",
       "         7.22656250e-02,  9.06024814e-03,  2.51130104e+01]),\n",
       " array([-6.19571594e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         8.69140625e-02,  1.30072325e-02,  1.94364980e+01]),\n",
       " array([-2.24681107e+02,  8.46134033e+01,  5.42307129e+01, ...,\n",
       "         3.51562500e-02,  2.12052175e-02,  1.29871875e+01]),\n",
       " array([-4.33180267e+02,  1.54702454e+02, -1.49000893e+01, ...,\n",
       "         1.48925781e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-5.22162781e+02,  3.24474602e+01,  5.34908152e+00, ...,\n",
       "         3.61328125e-02,  1.96078604e-02,  1.81663603e+01]),\n",
       " array([-5.06176849e+02,  4.39514084e+01,  2.28018456e+01, ...,\n",
       "         4.43359375e-01,  1.14887561e-02,  2.27802694e+01]),\n",
       " array([-5.64140259e+02,  7.79410172e+01,  1.10335140e+01, ...,\n",
       "         8.69140625e-02,  2.54132010e-02,  9.57651809e+00]),\n",
       " array([-2.12333588e+02,  8.34649963e+01,  3.45954208e+01, ...,\n",
       "         2.79296875e-01,  6.45462670e-03,  1.54451370e+01]),\n",
       " array([-4.15406921e+02,  4.80459785e+01, -2.32552738e+01, ...,\n",
       "         6.20117188e-02,  2.38023879e-02,  3.00817966e+01]),\n",
       " array([-5.25042114e+02,  6.28990078e+00, -2.19125786e+01, ...,\n",
       "         1.75781250e-02,  2.26311654e-02,  1.49979517e+01]),\n",
       " array([-2.87982269e+02, -2.25131340e+01, -2.61583786e+01, ...,\n",
       "         2.56835938e-01,  6.27334876e-03,  1.96148813e+01]),\n",
       " array([-5.77561462e+02,  3.11687241e+01,  2.54257870e+01, ...,\n",
       "         8.59375000e-02,  1.41425131e-02,  8.42891484e+00]),\n",
       " array([-5.39141907e+02,  1.19221077e+01,  1.00794582e+01, ...,\n",
       "         3.90625000e-02,  2.23559674e-02,  2.80899763e+01]),\n",
       " array([-2.71559418e+02,  1.79401264e+01, -1.61222572e+01, ...,\n",
       "         5.95703125e-02,  2.79151243e-02,  1.23045310e+01]),\n",
       " array([-2.14935593e+02,  3.74924088e+00,  6.17902422e+00, ...,\n",
       "         2.40234375e-01,  1.16592902e-02,  1.04809530e+01]),\n",
       " array([-2.37394485e+02,  1.61727886e+01,  1.78053150e+01, ...,\n",
       "         4.44335938e-02,  4.43977341e-02,  1.32469714e+01]),\n",
       " array([-3.72563263e+02,  6.36300964e+01,  9.21515083e+00, ...,\n",
       "         7.91015625e-02,  9.76588111e-03,  1.34223655e+01]),\n",
       " array([-6.16302063e+02,  3.28101082e+01,  2.65911698e+00, ...,\n",
       "         1.26953125e-01,  7.42948485e-03,  1.44374251e+01]),\n",
       " array([-4.54247406e+02, -1.09410896e+01, -5.87850189e+01, ...,\n",
       "         1.45996094e-01,  1.08151607e-02,  2.00595334e+01]),\n",
       " array([-2.30772675e+02,  1.21483170e+02,  2.34685535e+01, ...,\n",
       "         4.88281250e-03,  1.84172110e-02,  1.41733438e+01]),\n",
       " array([-5.80611084e+02,  8.87809753e+01,  3.08596840e+01, ...,\n",
       "         1.58691406e-01,  8.27864588e-03,  2.34573171e+01]),\n",
       " array([-3.08848419e+02, -1.20396652e+02,  1.18529453e+01, ...,\n",
       "         1.45019531e-01,  1.07390749e-02,  2.14375690e+01]),\n",
       " array([-5.77503540e+02,  5.21373367e+00, -1.75740681e+01, ...,\n",
       "         7.71484375e-02,  5.89116597e-03,  9.79063064e+00]),\n",
       " array([-9.81947708e+01,  1.09636215e+02, -1.55966830e+00, ...,\n",
       "         1.02539062e-02,  3.20658393e-02,  1.24913365e+01]),\n",
       " array([-5.76167053e+02,  4.06845627e+01,  1.08631706e+01, ...,\n",
       "         5.32226562e-02,  8.27289893e-03,  1.63605049e+01]),\n",
       " array([-9.52783661e+01,  9.00554810e+01, -5.58909416e-01, ...,\n",
       "         1.53808594e-01,  5.56502532e-03,  1.25259280e+01]),\n",
       " array([-6.09583069e+02,  0.00000000e+00,  0.00000000e+00, ...,\n",
       "         2.12890625e-01,  1.89839498e-02,  1.47575498e+01]),\n",
       " array([-2.98139862e+02, -8.89503937e+01,  1.02518444e+01, ...,\n",
       "         8.78906250e-02,  9.66035245e-03,  1.74519941e+01]),\n",
       " array([-3.94975281e+02,  1.41458801e+02,  4.71512299e+01, ...,\n",
       "         1.41601562e-02,  1.40249939e-02,  1.62735507e+01]),\n",
       " array([-1.79546616e+02,  6.36553268e+01, -8.02984009e+01, ...,\n",
       "         1.73339844e-01,  4.88385841e-03,  2.06619516e+01]),\n",
       " array([-4.19595093e+02,  1.02366013e+02,  8.81946182e+00, ...,\n",
       "         6.54296875e-02,  2.76340947e-02,  1.52613595e+01]),\n",
       " array([-1.94079620e+02,  1.49086266e+01, -4.39023743e+01, ...,\n",
       "         1.31835938e-01,  2.83286608e-02,  1.67268232e+01]),\n",
       " array([-2.66460968e+02,  1.03657043e+02, -4.25117416e+01, ...,\n",
       "         3.75976562e-02,  1.39411932e-02,  2.51948237e+01]),\n",
       " array([-1.97515106e+02,  1.78439148e+02,  8.76210403e+00, ...,\n",
       "         2.01660156e-01,  1.37209897e-02,  1.49848878e+01]),\n",
       " array([-6.15333923e+02,  8.55475545e-01,  7.60443449e-01, ...,\n",
       "         4.98046875e-02,  1.51160906e-02,  1.19729750e+01]),\n",
       " array([-6.13522217e+02,  1.82542877e+01,  1.50463829e+01, ...,\n",
       "         5.27343750e-02,  2.48250318e-02,  1.51770294e+01]),\n",
       " array([-5.11783203e+02,  7.93306732e+01,  4.38065262e+01, ...,\n",
       "         2.34375000e-02,  2.31996624e-02,  2.01258808e+01]),\n",
       " array([-3.02625824e+02,  1.06583359e+02,  3.97807007e+01, ...,\n",
       "         1.25488281e-01,  3.02432606e-02,  1.35143414e+01]),\n",
       " array([-1.49426224e+02,  1.17452377e+02, -3.53530579e+01, ...,\n",
       "         4.81445312e-01,  7.79947025e-03,  2.37603202e+01]),\n",
       " array([-5.11105865e+02,  6.43929672e+00, -9.46719933e+00, ...,\n",
       "         6.15234375e-02,  4.03804241e-03,  1.63724825e+01]),\n",
       " array([-8.04115540e+02, -1.77893519e-01, -2.08230686e+00, ...,\n",
       "         6.44531250e-02,  1.82596614e-02,  7.76281729e+00]),\n",
       " array([-2.30716019e+02,  1.07908623e+02,  3.31592255e+01, ...,\n",
       "         2.63671875e-02,  2.81991803e-02,  2.60955989e+01]),\n",
       " array([-3.06920105e+02,  1.40504654e+02,  1.07003880e+01, ...,\n",
       "         1.27929688e-01,  1.36056070e-02,  1.85453445e+01]),\n",
       " array([-4.87304077e+01,  4.85865364e+01, -4.00197258e+01, ...,\n",
       "         1.83105469e-01,  4.22899211e-03,  1.23187892e+01]),\n",
       " array([-5.99906799e+02,  8.21671524e+01,  2.59536972e+01, ...,\n",
       "         7.03125000e-02,  1.73223562e-02,  2.53322899e+01]),\n",
       " array([-4.78513794e+02,  1.09105015e+01,  4.85511494e+00, ...,\n",
       "         4.15039062e-02,  5.62006020e-03,  1.63083881e+01]),\n",
       " array([-2.05467468e+02,  1.36708176e+02, -1.11658783e+01, ...,\n",
       "         6.10351562e-02,  2.77174141e-02,  1.66221544e+01]),\n",
       " array([-5.08429352e+02,  4.16374283e+01,  1.67832012e+01, ...,\n",
       "         6.20117188e-02,  1.81225896e-02,  1.70100734e+01]),\n",
       " array([-1.45840012e+02, -7.36419373e+01,  8.23159866e+01, ...,\n",
       "         1.36718750e-02,  1.58701599e-02,  1.18854649e+01]),\n",
       " array([-4.11775757e+02,  6.80381927e+01, -3.73511124e+01, ...,\n",
       "         2.56835938e-01,  2.29659055e-02,  2.23294750e+01]),\n",
       " array([-5.31588562e+02,  8.86821747e-01,  7.85318255e-01, ...,\n",
       "         2.34375000e-02,  3.28308973e-03,  1.07292630e+01]),\n",
       " array([-1.26084305e+02,  7.56068726e+01,  2.84767866e+00, ...,\n",
       "         6.20117188e-02,  3.94357939e-02,  2.63452917e+01]),\n",
       " array([-3.35659027e+02,  1.43777618e+02, -8.42239857e+00, ...,\n",
       "         9.61914062e-02,  1.88302347e-02,  2.11464897e+01]),\n",
       " array([-1.56249649e+02,  1.73968002e+02, -9.20576172e+01, ...,\n",
       "         7.56835938e-02,  1.45951288e-02,  2.59964794e+01]),\n",
       " array([-2.68150116e+02,  1.02238937e+02,  5.62434959e+00, ...,\n",
       "         5.31738281e-01,  1.41048641e-02,  1.58088088e+01]),\n",
       " array([-8.99179688e+01,  1.16679657e+02, -3.77851677e+01, ...,\n",
       "         1.16699219e-01,  1.73497034e-02,  1.15830421e+01]),\n",
       " array([-3.08676270e+02, -1.89533386e+01, -3.87006807e+00, ...,\n",
       "         4.88281250e-02,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-3.94378174e+02,  9.60442657e+01,  1.28158627e+01, ...,\n",
       "         4.83398438e-02,  1.70821326e-02,  2.03577101e+01]),\n",
       " array([-4.77807251e+02,  3.43845444e+01,  1.63234901e+00, ...,\n",
       "         1.49902344e-01,  2.07116192e-03,  1.03307024e+01]),\n",
       " array([-5.94561279e+02,  6.57412109e+01,  1.25101051e+01, ...,\n",
       "         8.74023438e-02,  3.26578498e-02,  1.38312906e+01]),\n",
       " array([-5.69017212e+02,  8.34373016e+01,  1.60758610e+01, ...,\n",
       "         1.84082031e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " array([-1.37527618e+02,  7.31188049e+01, -1.85042381e+01, ...,\n",
       "         4.68750000e-02,  3.11771405e-02,  2.38001347e+01]),\n",
       " array([-3.75177795e+02, -2.13413391e+01, -4.01857605e+01, ...,\n",
       "         2.13378906e-01,  9.10250999e-03,  1.57136008e+01]),\n",
       " array([-2.94803864e+02,  1.48467926e+02, -1.89964828e+01, ...,\n",
       "         5.32226562e-02,  1.76829141e-02,  2.35499322e+01]),\n",
       " array([-3.29619598e+02,  5.61823368e-01,  4.43894958e+01, ...,\n",
       "         5.74218750e-01,  1.23265665e-02,  1.32037044e+01]),\n",
       " array([-3.80320251e+02,  3.35357208e+01, -1.54642654e+02, ...,\n",
       "         2.04589844e-01,  0.00000000e+00,  0.00000000e+00]),\n",
       " ...]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merged_features_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "2e5be546-a873-4fc9-be00-1578bef8fdd1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of merged_features_train: (2891, 3856)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of merged_features_train:\", np.array(merged_features_train).shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "36739907-b339-427b-b692-945b75069239",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_temp_df = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "13f5a524-ec8d-4fca-8a8d-32eff5ddcb3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract MFCC features for testing set\n",
    "test_temp_df['mfcc'] = Parallel(n_jobs=-1)(delayed(extract_mfcc)(audio_file) for audio_file in X_test)\n",
    "\n",
    "# Extract ZCR features for testing set\n",
    "test_temp_df['zcr'] = Parallel(n_jobs=-1)(delayed(extract_zcr)(audio_file) for audio_file in X_test)\n",
    "\n",
    "# Extract Jitter features for testing set\n",
    "test_temp_df['jitter'] = Parallel(n_jobs=-1)(delayed(calculate_jitter)(audio_file) for audio_file in X_test)\n",
    "\n",
    "# Replace NaN values in Jitter column with zero for testing set\n",
    "test_temp_df['jitter'] = test_temp_df['jitter'].fillna(0)\n",
    "\n",
    "# Extract Shimmer features for testing set\n",
    "test_temp_df['shimmer'] = Parallel(n_jobs=-1)(delayed(calculate_shimmer)(audio_file) for audio_file in X_test)\n",
    "\n",
    "# Replace NaN values in Shimmer column with zero for testing set\n",
    "test_temp_df['shimmer'] = test_temp_df['shimmer'].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "a15b091a-0437-4801-b618-1974f6b82aa1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "mfcc       0\n",
       "zcr        0\n",
       "jitter     0\n",
       "shimmer    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_temp_df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "3a27e221-8c65-4771-bdeb-c64193ccd11c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert testing dataframe to arrays\n",
    "mfcc_features_test = np.array(test_temp_df['mfcc'])\n",
    "zcr_features_test = np.array(test_temp_df['zcr'])\n",
    "jitter_features_test = np.array(test_temp_df['jitter'])\n",
    "shimmer_features_test = np.array(test_temp_df['shimmer'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "40c8f262-541b-4a10-9f5c-a04e21c149e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "merged_features_test = [merge_features(mfcc, zcr, jitter, shimmer) for mfcc, zcr, jitter, shimmer in zip(mfcc_features_test, zcr_features_test, jitter_features_test, shimmer_features_test)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "59bcec29-0960-42a2-8d5b-b37ed1f45848",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of merged_features_test: (1239, 3856)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of merged_features_test:\", np.array(merged_features_test).shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "76832657-5453-4373-b1a9-21779aa1dfd3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribution of 0s and 1s in y_train:\n",
      "1    1446\n",
      "0    1445\n",
      "Name: count, dtype: int64\n",
      "\n",
      "Distribution of 0s and 1s in y_test:\n",
      "0    620\n",
      "1    619\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "y_train_series = pd.Series(y_train)\n",
    "y_test_series = pd.Series(y_test)\n",
    "\n",
    "# Count the occurrences of each unique value in y_train\n",
    "train_distribution = y_train_series.value_counts()\n",
    "\n",
    "# Count the occurrences of each unique value in y_test\n",
    "test_distribution = y_test_series.value_counts()\n",
    "\n",
    "print(\"Distribution of 0s and 1s in y_train:\")\n",
    "print(train_distribution)\n",
    "\n",
    "print(\"\\nDistribution of 0s and 1s in y_test:\")\n",
    "print(test_distribution)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33fc6541-6107-4087-8830-cecbed96bbda",
   "metadata": {},
   "source": [
    "# Random Forest"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8821d5f-ba03-42c8-b51a-43b1808b919a",
   "metadata": {},
   "source": [
    "### Searching for best parameter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "25d4f7d8-2ee5-4321-9366-35b556609bc2",
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_rf = {\n",
    "    'n_estimators': [100, 300, 500, 800],\n",
    "    'max_depth': [5, 10, 15, 20],\n",
    "    'min_samples_split': [2, 5, 10, 20],\n",
    "    'min_samples_leaf': [1, 2, 4, 6, 10]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "e371e0b8-47a1-440f-909b-08585fb31146",
   "metadata": {},
   "outputs": [],
   "source": [
    "rf = RandomForestClassifier(random_state=42, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "ade6819e-9ca3-4206-8c81-3df92081c57c",
   "metadata": {},
   "outputs": [],
   "source": [
    "random_search_rf = RandomizedSearchCV(estimator=rf, param_distributions=param_grid_rf, n_iter=50, cv=5, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "e5f69d63-07a9-403c-a821-c02242ee4ef2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomizedSearchCV(cv=5,\n",
       "                   estimator=RandomForestClassifier(n_jobs=-1, random_state=42),\n",
       "                   n_iter=50,\n",
       "                   param_distributions={&#x27;max_depth&#x27;: [5, 10, 15, 20],\n",
       "                                        &#x27;min_samples_leaf&#x27;: [1, 2, 4, 6, 10],\n",
       "                                        &#x27;min_samples_split&#x27;: [2, 5, 10, 20],\n",
       "                                        &#x27;n_estimators&#x27;: [100, 300, 500, 800]},\n",
       "                   random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomizedSearchCV</label><div class=\"sk-toggleable__content\"><pre>RandomizedSearchCV(cv=5,\n",
       "                   estimator=RandomForestClassifier(n_jobs=-1, random_state=42),\n",
       "                   n_iter=50,\n",
       "                   param_distributions={&#x27;max_depth&#x27;: [5, 10, 15, 20],\n",
       "                                        &#x27;min_samples_leaf&#x27;: [1, 2, 4, 6, 10],\n",
       "                                        &#x27;min_samples_split&#x27;: [2, 5, 10, 20],\n",
       "                                        &#x27;n_estimators&#x27;: [100, 300, 500, 800]},\n",
       "                   random_state=42)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(n_jobs=-1, random_state=42)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(n_jobs=-1, random_state=42)</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomizedSearchCV(cv=5,\n",
       "                   estimator=RandomForestClassifier(n_jobs=-1, random_state=42),\n",
       "                   n_iter=50,\n",
       "                   param_distributions={'max_depth': [5, 10, 15, 20],\n",
       "                                        'min_samples_leaf': [1, 2, 4, 6, 10],\n",
       "                                        'min_samples_split': [2, 5, 10, 20],\n",
       "                                        'n_estimators': [100, 300, 500, 800]},\n",
       "                   random_state=42)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search_rf.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "27ea2cbd-f9bf-4327-a5fa-567fafdae5bf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters (Random Forest): {'n_estimators': 800, 'min_samples_split': 5, 'min_samples_leaf': 4, 'max_depth': 10}\n",
      "Best score (Random Forest): 65.6508%\n"
     ]
    }
   ],
   "source": [
    "best_params_rf = random_search_rf.best_params_\n",
    "best_model_rf = random_search_rf.best_estimator_\n",
    "\n",
    "print(\"Best parameters (Random Forest):\", best_params_rf)\n",
    "\n",
    "print(f\"Best score (Random Forest): {random_search_rf.best_score_*100:.4f}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b85adc9b-2bb3-4895-9457-9a2ebb3cb7f3",
   "metadata": {},
   "source": [
    "#### Evaluate the best model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "34a68d48-fcfa-44d6-be57-c8989c97bca6",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve, auc\n",
    "def plot_roc(best_model, model_name):\n",
    "    predicted_probabilities = best_model.predict_proba(merged_features_test)[:, 1]\n",
    "    fpr, tpr, _ = roc_curve(y_test, predicted_probabilities)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    \n",
    "    # Plot ROC curve\n",
    "    plt.figure()\n",
    "    plt.plot(fpr, tpr, color='darkorange', lw=2, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "    plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')\n",
    "    plt.xlim([0.0, 1.0])\n",
    "    plt.ylim([0.0, 1.05])\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title(f'Prolongation - {model_name} - ROC Curve')\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    plt.savefig(f'ROC_Prolongation_{model_name}.png')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "99503401-eac5-44f2-b92a-4c77a421f8ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_rf = best_model_rf.predict(merged_features_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "dab91af6-7ed8-46d4-8e82-d921cf5c4aa0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6577885391444713"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test,y_pred_rf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "ca5eb1cc-5fb1-47e8-b325-1af2fee84582",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.64      0.71      0.68       620\n",
      "           1       0.68      0.60      0.64       619\n",
      "\n",
      "    accuracy                           0.66      1239\n",
      "   macro avg       0.66      0.66      0.66      1239\n",
      "weighted avg       0.66      0.66      0.66      1239\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_rf))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "14e27f2e-ebb7-4bc5-a488-1d620df49571",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomForestClassifier(max_depth=10, min_samples_leaf=4, min_samples_split=5,\n",
       "                       n_estimators=800, n_jobs=-1)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" checked><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(max_depth=10, min_samples_leaf=4, min_samples_split=5,\n",
       "                       n_estimators=800, n_jobs=-1)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestClassifier(max_depth=10, min_samples_leaf=4, min_samples_split=5,\n",
       "                       n_estimators=800, n_jobs=-1)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_rf = RandomForestClassifier(n_jobs=-1, n_estimators=800, min_samples_split=5, min_samples_leaf=4, max_depth=10)\n",
    "best_model_rf.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "d62c4e6c-b927-4047-9158-9b2d2ac1f832",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(best_model_rf, 'Random Forest')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "5eaa7c32-60ed-49ef-b040-48cfc83916bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_rf = pd.DataFrame([X_test, y_pred_rf, y_test]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "43810d65-b43d-4a6f-aca8-ee5e57cfe978",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_rf.columns = ['audio', 'y_pred', 'y_test']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "e3a95d14-d1c2-43c5-b616-8444936d4e91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>audio</th>\n",
       "      <th>y_pred</th>\n",
       "      <th>y_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1229</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1232</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1233</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1236</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1238</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>442 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  audio y_pred y_test\n",
       "1     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "4     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "5     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "12    C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "16    C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "...                                                 ...    ...    ...\n",
       "1229  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "1232  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "1233  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "1236  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "1238  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "\n",
       "[442 rows x 3 columns]"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "verifyer_rf[(verifyer_rf['y_pred']==0) & (verifyer_rf['y_test']==0)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "f7489d68-410d-4a42-9a98-c70c6af49845",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>audio</th>\n",
       "      <th>y_pred</th>\n",
       "      <th>y_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1228</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1230</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1231</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1235</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1237</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>373 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  audio y_pred y_test\n",
       "0     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "2     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "7     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "9     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "14    C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "...                                                 ...    ...    ...\n",
       "1228  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1230  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1231  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1235  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1237  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "\n",
       "[373 rows x 3 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "verifyer_rf[(verifyer_rf['y_pred']==1) & (verifyer_rf['y_test']==1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "a56c11e7-ab31-4614-a462-8a64c3b17307",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>audio</th>\n",
       "      <th>y_pred</th>\n",
       "      <th>y_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1234</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1235</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1236</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1237</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1238</th>\n",
       "      <td>C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1239 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  audio y_pred y_test\n",
       "0     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "2     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "3     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      1\n",
       "4     C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "...                                                 ...    ...    ...\n",
       "1234  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      0\n",
       "1235  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1236  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "1237  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      1      1\n",
       "1238  C:\\Users\\sakshi\\Desktop\\SMLProject\\DatasetExtr...      0      0\n",
       "\n",
       "[1239 rows x 3 columns]"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "verifyer_rf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "23d1e458-5a3f-4209-8af3-c483afcb7f8a",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_rf.to_csv(\"predicted_randomforest_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "cc4d895c-436e-41ee-b46c-d37d8b20c5e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "file_path = \"randomforest_Prolongation.pkl\"\n",
    "\n",
    "with open(file_path, 'wb') as file:\n",
    "    pickle.dump(best_model_rf, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "8b1b5ca6-f119-4185-a2e7-ee90d05d9655",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_rf = pd.DataFrame([X_test, y_test, y_pred_rf]).T\n",
    "test_rf.columns = ['X_test', 'y_test', 'y_pred']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "e6c3db53-48c7-43b5-8dcb-595c4fe58dbd",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_rf.to_csv(\"test_rf_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e60af24-2263-4b20-866a-68cf60dcc54e",
   "metadata": {},
   "source": [
    "# Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "2cade8e4-721c-47ca-a1eb-a03bd435cc25",
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_dt = {\n",
    "    'max_depth': [5, 10, 15, 20],\n",
    "    'min_samples_split': [2, 5, 10, 20],\n",
    "    'min_samples_leaf': [1, 2, 4, 6, 10]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "2785778b-ab6a-482a-b4ca-3403b8b7a81c",
   "metadata": {},
   "outputs": [],
   "source": [
    "dt = DecisionTreeClassifier(random_state=42)\n",
    "random_search_dt = RandomizedSearchCV(estimator=dt, param_distributions=param_grid_dt, n_iter=50, cv=5, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "b7221af6-7dc9-4e21-9cbb-a0a476372806",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomizedSearchCV(cv=5, estimator=DecisionTreeClassifier(random_state=42),\n",
       "                   n_iter=50,\n",
       "                   param_distributions={&#x27;max_depth&#x27;: [5, 10, 15, 20],\n",
       "                                        &#x27;min_samples_leaf&#x27;: [1, 2, 4, 6, 10],\n",
       "                                        &#x27;min_samples_split&#x27;: [2, 5, 10, 20]},\n",
       "                   random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomizedSearchCV</label><div class=\"sk-toggleable__content\"><pre>RandomizedSearchCV(cv=5, estimator=DecisionTreeClassifier(random_state=42),\n",
       "                   n_iter=50,\n",
       "                   param_distributions={&#x27;max_depth&#x27;: [5, 10, 15, 20],\n",
       "                                        &#x27;min_samples_leaf&#x27;: [1, 2, 4, 6, 10],\n",
       "                                        &#x27;min_samples_split&#x27;: [2, 5, 10, 20]},\n",
       "                   random_state=42)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: DecisionTreeClassifier</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeClassifier(random_state=42)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">DecisionTreeClassifier</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeClassifier(random_state=42)</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomizedSearchCV(cv=5, estimator=DecisionTreeClassifier(random_state=42),\n",
       "                   n_iter=50,\n",
       "                   param_distributions={'max_depth': [5, 10, 15, 20],\n",
       "                                        'min_samples_leaf': [1, 2, 4, 6, 10],\n",
       "                                        'min_samples_split': [2, 5, 10, 20]},\n",
       "                   random_state=42)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search_dt.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "ce8e5c28-d88c-41b8-9d9a-2a17951c0554",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters (Decision Tree): {'min_samples_split': 2, 'min_samples_leaf': 1, 'max_depth': 5}\n",
      "Best score (Decision Tree): 58.0423%\n"
     ]
    }
   ],
   "source": [
    "best_params_dt = random_search_dt.best_params_\n",
    "best_model_dt = random_search_dt.best_estimator_\n",
    "\n",
    "print(\"Best parameters (Decision Tree):\", best_params_dt)\n",
    "\n",
    "print(f\"Best score (Decision Tree): {random_search_dt.best_score_*100:.4f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "c49c9f1f-7c4e-4776-8ed9-37dc1cb379ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_dt = best_model_dt.predict(merged_features_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "86b0e8c2-d4a1-4a77-8805-8355ae330cb8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5754640839386602"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, y_pred_dt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "2acd500d-0c5c-4179-b995-023af746b3ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.56      0.69      0.62       620\n",
      "           1       0.60      0.46      0.52       619\n",
      "\n",
      "    accuracy                           0.58      1239\n",
      "   macro avg       0.58      0.58      0.57      1239\n",
      "weighted avg       0.58      0.58      0.57      1239\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_dt))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "d07bb2e8-87c1-4307-8c4d-e1e57694c61d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-4 {color: black;background-color: white;}#sk-container-id-4 pre{padding: 0;}#sk-container-id-4 div.sk-toggleable {background-color: white;}#sk-container-id-4 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-4 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-4 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-4 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-4 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-4 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-4 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-4 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-4 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-4 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-4 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-4 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-4 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-4 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-4 div.sk-item {position: relative;z-index: 1;}#sk-container-id-4 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-4 div.sk-item::before, #sk-container-id-4 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-4 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-4 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-4 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-4 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-4 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-4 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-4 div.sk-label-container {text-align: center;}#sk-container-id-4 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-4 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-4\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>DecisionTreeClassifier(max_depth=5)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" checked><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">DecisionTreeClassifier</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeClassifier(max_depth=5)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "DecisionTreeClassifier(max_depth=5)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_dt = DecisionTreeClassifier(min_samples_split=2, min_samples_leaf=1, max_depth=5)\n",
    "best_model_dt.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "a7b6277c-d2ca-4207-9c3d-7dcb1b090b35",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(best_model_dt,'Decision Tree')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "42b0dea2-9d90-44a9-af96-55cd38e6f606",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = \"decisiontree_Prolongation.pkl\"\n",
    "\n",
    "with open(file_path, 'wb') as file:\n",
    "    pickle.dump(best_model_dt, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "7dd47cb8-42ba-42cb-bfcc-95be747c9c2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_dt = pd.DataFrame([X_test, y_pred_dt, y_test]).T\n",
    "verifyer_dt.columns = ['audio', 'y_pred', 'y_test']\n",
    "verifyer_dt.to_csv(\"predicted_dt_prolongation.csv\")\n",
    "test_dt = pd.DataFrame([X_test, y_test, y_pred_dt]).T\n",
    "test_dt.columns = ['X_test', 'y_test', 'y_pred']\n",
    "test_dt.to_csv(\"test_dt_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c7b67ccf-66e0-41a1-bf08-423f87189750",
   "metadata": {},
   "source": [
    "# KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "0ddb7fcc-b0ec-4926-b15d-58220290e070",
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_knn = {\n",
    "    'n_neighbors': np.arange(1, 22, 2),\n",
    "    'weights': ['uniform', 'distance'],\n",
    "    'algorithm': ['auto', 'ball_tree', 'kd_tree', 'brute'],\n",
    "    'p': [1, 2]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "5598004c-f634-4224-bc7e-8cbf08e8b5d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "knn = KNeighborsClassifier(n_jobs=-1)\n",
    "random_search_knn = RandomizedSearchCV(estimator=knn, param_distributions=param_grid_knn, n_iter=50, cv=5, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "c36ede34-331b-4f39-9556-fb2adf899322",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-5 {color: black;background-color: white;}#sk-container-id-5 pre{padding: 0;}#sk-container-id-5 div.sk-toggleable {background-color: white;}#sk-container-id-5 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-5 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-5 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-5 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-5 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-5 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-5 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-5 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-5 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-5 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-5 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-5 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-5 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-5 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-5 div.sk-item {position: relative;z-index: 1;}#sk-container-id-5 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-5 div.sk-item::before, #sk-container-id-5 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-5 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-5 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-5 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-5 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-5 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-5 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-5 div.sk-label-container {text-align: center;}#sk-container-id-5 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-5 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-5\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomizedSearchCV(cv=5, estimator=KNeighborsClassifier(n_jobs=-1), n_iter=50,\n",
       "                   param_distributions={&#x27;algorithm&#x27;: [&#x27;auto&#x27;, &#x27;ball_tree&#x27;,\n",
       "                                                      &#x27;kd_tree&#x27;, &#x27;brute&#x27;],\n",
       "                                        &#x27;n_neighbors&#x27;: array([ 1,  3,  5,  7,  9, 11, 13, 15, 17, 19, 21]),\n",
       "                                        &#x27;p&#x27;: [1, 2],\n",
       "                                        &#x27;weights&#x27;: [&#x27;uniform&#x27;, &#x27;distance&#x27;]},\n",
       "                   random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomizedSearchCV</label><div class=\"sk-toggleable__content\"><pre>RandomizedSearchCV(cv=5, estimator=KNeighborsClassifier(n_jobs=-1), n_iter=50,\n",
       "                   param_distributions={&#x27;algorithm&#x27;: [&#x27;auto&#x27;, &#x27;ball_tree&#x27;,\n",
       "                                                      &#x27;kd_tree&#x27;, &#x27;brute&#x27;],\n",
       "                                        &#x27;n_neighbors&#x27;: array([ 1,  3,  5,  7,  9, 11, 13, 15, 17, 19, 21]),\n",
       "                                        &#x27;p&#x27;: [1, 2],\n",
       "                                        &#x27;weights&#x27;: [&#x27;uniform&#x27;, &#x27;distance&#x27;]},\n",
       "                   random_state=42)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-10\" type=\"checkbox\" ><label for=\"sk-estimator-id-10\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: KNeighborsClassifier</label><div class=\"sk-toggleable__content\"><pre>KNeighborsClassifier(n_jobs=-1)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-11\" type=\"checkbox\" ><label for=\"sk-estimator-id-11\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">KNeighborsClassifier</label><div class=\"sk-toggleable__content\"><pre>KNeighborsClassifier(n_jobs=-1)</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomizedSearchCV(cv=5, estimator=KNeighborsClassifier(n_jobs=-1), n_iter=50,\n",
       "                   param_distributions={'algorithm': ['auto', 'ball_tree',\n",
       "                                                      'kd_tree', 'brute'],\n",
       "                                        'n_neighbors': array([ 1,  3,  5,  7,  9, 11, 13, 15, 17, 19, 21]),\n",
       "                                        'p': [1, 2],\n",
       "                                        'weights': ['uniform', 'distance']},\n",
       "                   random_state=42)"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search_knn.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "f947deac-f696-4f5f-967e-0b021986c93c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters (KNN): {'weights': 'distance', 'p': 1, 'n_neighbors': 21, 'algorithm': 'auto'}\n",
      "Best score (KNN): 62.4356%\n"
     ]
    }
   ],
   "source": [
    "best_params_knn = random_search_knn.best_params_\n",
    "best_model_knn = random_search_knn.best_estimator_\n",
    "\n",
    "print(\"Best parameters (KNN):\", best_params_knn)\n",
    "\n",
    "print(f\"Best score (KNN): {random_search_knn.best_score_*100:.4f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "1fada41c-d6cd-4711-9630-6e659e86536c",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_knn = best_model_knn.predict(merged_features_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "28195ac2-1106-4b1c-97fc-e208367c52eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6045197740112994"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, y_pred_knn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "6d3c689a-094a-478e-8df6-9c72d6767b89",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.61      0.59      0.60       620\n",
      "           1       0.60      0.62      0.61       619\n",
      "\n",
      "    accuracy                           0.60      1239\n",
      "   macro avg       0.60      0.60      0.60      1239\n",
      "weighted avg       0.60      0.60      0.60      1239\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_knn))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "ce36c200-b47e-4766-aa9b-f38234063e5e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-6 {color: black;background-color: white;}#sk-container-id-6 pre{padding: 0;}#sk-container-id-6 div.sk-toggleable {background-color: white;}#sk-container-id-6 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-6 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-6 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-6 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-6 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-6 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-6 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-6 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-6 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-6 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-6 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-6 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-6 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-6 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-6 div.sk-item {position: relative;z-index: 1;}#sk-container-id-6 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-6 div.sk-item::before, #sk-container-id-6 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-6 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-6 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-6 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-6 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-6 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-6 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-6 div.sk-label-container {text-align: center;}#sk-container-id-6 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-6 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-6\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>KNeighborsClassifier(n_jobs=-1, n_neighbors=21, p=1, weights=&#x27;distance&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-12\" type=\"checkbox\" checked><label for=\"sk-estimator-id-12\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">KNeighborsClassifier</label><div class=\"sk-toggleable__content\"><pre>KNeighborsClassifier(n_jobs=-1, n_neighbors=21, p=1, weights=&#x27;distance&#x27;)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "KNeighborsClassifier(n_jobs=-1, n_neighbors=21, p=1, weights='distance')"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_knn = KNeighborsClassifier(n_jobs=-1, weights='distance', p=1, n_neighbors=21, algorithm='auto')\n",
    "best_model_knn.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "2b7b8e6b-82c0-4c64-8d89-efe3a4928110",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(best_model_knn, 'KNN')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "94e52ea6-7b83-4be9-9143-e07bccba91aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = \"knn_Prolongation.pkl\"\n",
    "\n",
    "with open(file_path, 'wb') as file:\n",
    "    pickle.dump(best_model_knn, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "5e166c9a-ef9a-46e1-bc7e-6f9fc1ab80e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_knn = pd.DataFrame([X_test, y_pred_knn, y_test]).T\n",
    "verifyer_knn.columns = ['audio', 'y_pred', 'y_test']\n",
    "verifyer_knn.to_csv(\"predicted_knn_prolongation.csv\")\n",
    "test_knn = pd.DataFrame([X_test, y_test, y_pred_knn]).T\n",
    "test_knn.columns = ['X_test', 'y_test', 'y_pred']\n",
    "test_knn.to_csv(\"test_knn_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6720fded-4777-45bc-9be9-3a0f489f7fd2",
   "metadata": {},
   "source": [
    "# SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "27b031f3-d46e-4295-ae89-7c38cb484355",
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_svc = {\n",
    "    'C': [0.1, 1, 10, 100],\n",
    "    'kernel': ['linear', 'poly', 'rbf', 'sigmoid'],\n",
    "    'gamma': ['scale', 'auto']\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "d0d128ca-f6af-4470-89d7-36475e793350",
   "metadata": {},
   "outputs": [],
   "source": [
    "svc = SVC()\n",
    "random_search_svc = RandomizedSearchCV(estimator=svc, param_distributions=param_grid_svc, n_iter=32, cv=5, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "11b9e384-4351-4dec-a10d-e1ab06774279",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-7 {color: black;background-color: white;}#sk-container-id-7 pre{padding: 0;}#sk-container-id-7 div.sk-toggleable {background-color: white;}#sk-container-id-7 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-7 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-7 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-7 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-7 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-7 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-7 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-7 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-7 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-7 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-7 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-7 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-7 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-7 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-7 div.sk-item {position: relative;z-index: 1;}#sk-container-id-7 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-7 div.sk-item::before, #sk-container-id-7 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-7 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-7 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-7 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-7 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-7 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-7 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-7 div.sk-label-container {text-align: center;}#sk-container-id-7 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-7 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-7\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomizedSearchCV(cv=5, estimator=SVC(), n_iter=32,\n",
       "                   param_distributions={&#x27;C&#x27;: [0.1, 1, 10, 100],\n",
       "                                        &#x27;gamma&#x27;: [&#x27;scale&#x27;, &#x27;auto&#x27;],\n",
       "                                        &#x27;kernel&#x27;: [&#x27;linear&#x27;, &#x27;poly&#x27;, &#x27;rbf&#x27;,\n",
       "                                                   &#x27;sigmoid&#x27;]},\n",
       "                   random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-13\" type=\"checkbox\" ><label for=\"sk-estimator-id-13\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomizedSearchCV</label><div class=\"sk-toggleable__content\"><pre>RandomizedSearchCV(cv=5, estimator=SVC(), n_iter=32,\n",
       "                   param_distributions={&#x27;C&#x27;: [0.1, 1, 10, 100],\n",
       "                                        &#x27;gamma&#x27;: [&#x27;scale&#x27;, &#x27;auto&#x27;],\n",
       "                                        &#x27;kernel&#x27;: [&#x27;linear&#x27;, &#x27;poly&#x27;, &#x27;rbf&#x27;,\n",
       "                                                   &#x27;sigmoid&#x27;]},\n",
       "                   random_state=42)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-14\" type=\"checkbox\" ><label for=\"sk-estimator-id-14\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: SVC</label><div class=\"sk-toggleable__content\"><pre>SVC()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-15\" type=\"checkbox\" ><label for=\"sk-estimator-id-15\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomizedSearchCV(cv=5, estimator=SVC(), n_iter=32,\n",
       "                   param_distributions={'C': [0.1, 1, 10, 100],\n",
       "                                        'gamma': ['scale', 'auto'],\n",
       "                                        'kernel': ['linear', 'poly', 'rbf',\n",
       "                                                   'sigmoid']},\n",
       "                   random_state=42)"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search_svc.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "a04d2236-b8c6-4d04-af0c-74f19f0cf8fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters (SVC): {'kernel': 'rbf', 'gamma': 'scale', 'C': 10}\n",
      "Best score (SVC): 63.5411%\n"
     ]
    }
   ],
   "source": [
    "best_params_svc = random_search_svc.best_params_\n",
    "best_model_svc = random_search_svc.best_estimator_\n",
    "\n",
    "print(\"Best parameters (SVC):\", best_params_svc)\n",
    "\n",
    "print(f\"Best score (SVC): {random_search_svc.best_score_*100:.4f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "e0e77b7d-5b62-4369-94f5-f76d17d29493",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_svc = best_model_svc.predict(merged_features_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "79dc8559-b7d4-42ef-b9b4-009bc55a8057",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6287328490718321"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, y_pred_svc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "129e8ef7-f866-476f-b5e9-d826fe9e16a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.63      0.61      0.62       620\n",
      "           1       0.62      0.65      0.64       619\n",
      "\n",
      "    accuracy                           0.63      1239\n",
      "   macro avg       0.63      0.63      0.63      1239\n",
      "weighted avg       0.63      0.63      0.63      1239\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_svc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "787b14eb-2f4c-44f8-b8f1-782e3385763f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-8 {color: black;background-color: white;}#sk-container-id-8 pre{padding: 0;}#sk-container-id-8 div.sk-toggleable {background-color: white;}#sk-container-id-8 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-8 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-8 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-8 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-8 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-8 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-8 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-8 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-8 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-8 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-8 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-8 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-8 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-8 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-8 div.sk-item {position: relative;z-index: 1;}#sk-container-id-8 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-8 div.sk-item::before, #sk-container-id-8 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-8 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-8 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-8 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-8 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-8 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-8 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-8 div.sk-label-container {text-align: center;}#sk-container-id-8 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-8 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-8\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SVC(C=10, probability=True)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-16\" type=\"checkbox\" checked><label for=\"sk-estimator-id-16\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC(C=10, probability=True)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SVC(C=10, probability=True)"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_svc = SVC(kernel='rbf', gamma='scale', C=10, probability=True)\n",
    "best_model_svc.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "36c9bc2e-4ebe-43fd-93b2-6eff43d733fb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(best_model_svc, 'SVC')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "dbe9eb15-91f3-4a9e-a3ff-16b5b0c95246",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = \"svc_Prolongation.pkl\"\n",
    "\n",
    "with open(file_path, 'wb') as file:\n",
    "    pickle.dump(best_model_svc, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "d9839878-6b67-4aee-9a5d-aa26ad30513f",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_svc = pd.DataFrame([X_test, y_pred_svc, y_test]).T\n",
    "verifyer_svc.columns = ['audio', 'y_pred', 'y_test']\n",
    "verifyer_svc.to_csv(\"predicted_svc_prolongation.csv\")\n",
    "test_svc = pd.DataFrame([X_test, y_test, y_pred_svc]).T\n",
    "test_svc.columns = ['X_test', 'y_test', 'y_pred']\n",
    "test_svc.to_csv(\"test_svc_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d8439fcd-1278-4f65-beba-2908ec74b2ef",
   "metadata": {},
   "source": [
    "# Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "135805b1-990b-4e46-ab0b-e5a1dbdf5e65",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "nb = GaussianNB()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "42b4d2de-a734-40de-a7ab-766d6ddd5792",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Perform cross-validation\n",
    "cv_results = cross_validate(nb, merged_features_train, y_train, cv=10,\n",
    "                            scoring=['accuracy', 'precision', 'recall', 'f1'],\n",
    "                            return_estimator=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "7d06359b-5a58-4a8e-8f9f-e21a33366a3f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'fit_time': array([0.186095  , 0.18709826, 0.1823802 , 0.20356393, 0.2091043 ,\n",
       "        0.18766141, 0.19471908, 0.18765688, 0.16528726, 0.1896615 ]),\n",
       " 'score_time': array([0.04569483, 0.03764153, 0.03124452, 0.03124928, 0.02212763,\n",
       "        0.05540013, 0.03118515, 0.03118014, 0.03623939, 0.03169799]),\n",
       " 'estimator': [GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB(),\n",
       "  GaussianNB()],\n",
       " 'test_accuracy': array([0.55517241, 0.61591696, 0.64705882, 0.57093426, 0.53979239,\n",
       "        0.60553633, 0.58823529, 0.56747405, 0.53979239, 0.56401384]),\n",
       " 'test_precision': array([0.55970149, 0.61038961, 0.64827586, 0.57342657, 0.54545455,\n",
       "        0.624     , 0.58278146, 0.56737589, 0.54471545, 0.56617647]),\n",
       " 'test_recall': array([0.51724138, 0.64827586, 0.64827586, 0.56551724, 0.49655172,\n",
       "        0.53793103, 0.61111111, 0.55555556, 0.46527778, 0.53472222]),\n",
       " 'test_f1': array([0.53763441, 0.62876254, 0.64827586, 0.56944444, 0.5198556 ,\n",
       "        0.57777778, 0.59661017, 0.56140351, 0.50187266, 0.55      ])}"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "5bdfdc71-ebdd-4cff-9af7-627d0290cea5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract the best model based on accuracy\n",
    "best_model_index = cv_results['test_accuracy'].argmax()\n",
    "best_model_nb = cv_results['estimator'][best_model_index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "ce8fad7e-02b2-4541-b4ab-b1ba15406038",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "0fc4583c-c5af-425d-a272-9ad15c158b41",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-9 {color: black;background-color: white;}#sk-container-id-9 pre{padding: 0;}#sk-container-id-9 div.sk-toggleable {background-color: white;}#sk-container-id-9 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-9 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-9 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-9 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-9 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-9 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-9 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-9 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-9 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-9 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-9 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-9 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-9 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-9 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-9 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-9 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-9 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-9 div.sk-item {position: relative;z-index: 1;}#sk-container-id-9 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-9 div.sk-item::before, #sk-container-id-9 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-9 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-9 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-9 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-9 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-9 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-9 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-9 div.sk-label-container {text-align: center;}#sk-container-id-9 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-9 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-9\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GaussianNB()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-17\" type=\"checkbox\" checked><label for=\"sk-estimator-id-17\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GaussianNB</label><div class=\"sk-toggleable__content\"><pre>GaussianNB()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "GaussianNB()"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_nb.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "20f9a8fc-d945-46af-808b-8d6aa0bcb524",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_nb = best_model_nb.predict(merged_features_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "91c3b5b0-3ee0-43cc-ae8b-43b56c69d747",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5980629539951574"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, y_pred_nb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "9d758479-bcef-4735-aff1-be7822937f4b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.60      0.61      0.60       620\n",
      "           1       0.60      0.58      0.59       619\n",
      "\n",
      "    accuracy                           0.60      1239\n",
      "   macro avg       0.60      0.60      0.60      1239\n",
      "weighted avg       0.60      0.60      0.60      1239\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_nb))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "18ed6f85-baaa-49e4-9b5c-ac912edecd09",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(best_model_nb, 'Naive Bayes')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "1aef7d58-b749-452e-9894-f6875be29e8a",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = \"naivebayes_Prolongation.pkl\"\n",
    "\n",
    "with open(file_path, 'wb') as file:\n",
    "    pickle.dump(best_model_nb, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "d877df6d-7862-4f78-8b5c-fdc278af1ca4",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_nb = pd.DataFrame([X_test, y_pred_nb, y_test]).T\n",
    "verifyer_nb.columns = ['audio', 'y_pred', 'y_test']\n",
    "verifyer_nb.to_csv(\"predicted_nb_prolongation.csv\")\n",
    "test_nb = pd.DataFrame([X_test, y_test, y_pred_nb]).T\n",
    "test_nb.columns = ['X_test', 'y_test', 'y_pred']\n",
    "test_nb.to_csv(\"test_nb_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8f75704-68f5-4b70-bf7d-7a299c625bfe",
   "metadata": {},
   "source": [
    "# Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "93f9e17d-6470-4f4d-9077-e0fd7bc83e48",
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = LogisticRegression(n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "af7bada6-0224-4ff2-b8f4-3e0d189a2759",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Perform cross-validation\n",
    "cv_results_1= cross_validate(lr, merged_features_train, y_train, cv=10,\n",
    "                            scoring=['accuracy', 'precision', 'recall', 'f1'],\n",
    "                            return_estimator=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "5bde6aa2-8784-4816-a243-af64763d625f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'fit_time': array([7.28140163, 2.78112078, 2.64065504, 2.43764019, 2.6684556 ,\n",
       "        2.76599073, 2.7383852 , 2.70932627, 2.41272354, 2.34076619]),\n",
       " 'score_time': array([0.01552272, 0.01563239, 0.        , 0.01567268, 0.01603532,\n",
       "        0.00262022, 0.01201534, 0.0100522 , 0.0156424 , 0.00652623]),\n",
       " 'estimator': [LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1),\n",
       "  LogisticRegression(n_jobs=-1)],\n",
       " 'test_accuracy': array([0.55517241, 0.5916955 , 0.56055363, 0.53979239, 0.56401384,\n",
       "        0.51903114, 0.57093426, 0.57093426, 0.5432526 , 0.57093426]),\n",
       " 'test_precision': array([0.55633803, 0.6       , 0.55487805, 0.54347826, 0.55900621,\n",
       "        0.52205882, 0.56097561, 0.57142857, 0.54      , 0.57042254]),\n",
       " 'test_recall': array([0.54482759, 0.55862069, 0.62758621, 0.51724138, 0.62068966,\n",
       "        0.48965517, 0.63888889, 0.55555556, 0.5625    , 0.5625    ]),\n",
       " 'test_f1': array([0.55052265, 0.57857143, 0.58899676, 0.53003534, 0.58823529,\n",
       "        0.50533808, 0.5974026 , 0.56338028, 0.55102041, 0.56643357])}"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "92625aa0-c90b-4e32-94e8-1f88b8b4295c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract the best model based on accuracy\n",
    "best_model_index_1 = cv_results_1['test_accuracy'].argmax()\n",
    "best_model_lr = cv_results_1['estimator'][best_model_index_1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "a960c9fa-6de9-44b1-9c2b-9c5ec470edf8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_index_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "1b54454f-10c5-4670-a33d-4c1caf5e32d8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-10 {color: black;background-color: white;}#sk-container-id-10 pre{padding: 0;}#sk-container-id-10 div.sk-toggleable {background-color: white;}#sk-container-id-10 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-10 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-10 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-10 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-10 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-10 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-10 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-10 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-10 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-10 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-10 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-10 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-10 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-10 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-10 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-10 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-10 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-10 div.sk-item {position: relative;z-index: 1;}#sk-container-id-10 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-10 div.sk-item::before, #sk-container-id-10 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-10 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-10 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-10 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-10 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-10 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-10 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-10 div.sk-label-container {text-align: center;}#sk-container-id-10 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-10 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-10\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LogisticRegression(n_jobs=-1)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-18\" type=\"checkbox\" checked><label for=\"sk-estimator-id-18\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LogisticRegression</label><div class=\"sk-toggleable__content\"><pre>LogisticRegression(n_jobs=-1)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "LogisticRegression(n_jobs=-1)"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_lr.fit(merged_features_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "8fe6f2fe-3136-4b6f-8f46-b6f983dbbbeb",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_lr = best_model_lr.predict(merged_features_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "0ce6b2cf-0929-4b0b-b9ca-ec4e430c6e34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5504439063761097"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, y_pred_lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "fb7c26d5-cbcc-42d1-ae03-4ce66ec3ca8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(best_model_lr, 'Logistic Regression')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "a2581e0c-0f91-4375-ad89-79bdcd334761",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path = \"LogisticRegression_Prolongation.pkl\"\n",
    "\n",
    "with open(file_path, 'wb') as file:\n",
    "    pickle.dump(best_model_nb, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "96f86655-0f96-4a20-aa58-738d8e1b0408",
   "metadata": {},
   "outputs": [],
   "source": [
    "verifyer_lr = pd.DataFrame([X_test, y_pred_lr, y_test]).T\n",
    "verifyer_lr.columns = ['audio', 'y_pred', 'y_test']\n",
    "verifyer_lr.to_csv(\"predicted_lr_prolongation.csv\")\n",
    "test_lr = pd.DataFrame([X_test, y_test, y_pred_lr]).T\n",
    "test_lr.columns = ['X_test', 'y_test', 'y_pred']\n",
    "test_lr.to_csv(\"test_lr_prolongation.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e305e9a4-1e2d-4b26-8f95-a430dc134cec",
   "metadata": {},
   "source": [
    "# Comparitive Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "e580741f-6df3-43fd-b8fd-c762e41b6258",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "fac1d008-91e3-44d3-ad34-e8be5ea7f163",
   "metadata": {},
   "outputs": [],
   "source": [
    "models = ['Random Forest', 'Decision Tree', 'KNN', 'SVC', 'Naive Bayes', 'Logistic Regression']\n",
    "accuracy_scores = []\n",
    "precision_scores = []\n",
    "recall_scores = []\n",
    "f1_scores = []\n",
    "best_models = [best_model_rf, best_model_dt, best_model_knn, best_model_svc, best_model_nb, best_model_lr]\n",
    "y_preds = [y_pred_rf, y_pred_dt, y_pred_knn, y_pred_svc, y_pred_nb, y_pred_lr]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "5524e7a9-372c-4527-9094-df6faed4debf",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to calculate evaluation metrics\n",
    "def calculate_metrics(y_true, y_pred):\n",
    "    accuracy = accuracy_score(y_true, y_pred)\n",
    "    precision = precision_score(y_true, y_pred)\n",
    "    recall = recall_score(y_true, y_pred)\n",
    "    f1 = f1_score(y_true, y_pred)\n",
    "    return accuracy, precision, recall, f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "3e4eedba-4980-419c-ba62-b5132e68a802",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate evaluation metrics for each model\n",
    "for model, y_pred in zip(best_models, y_preds):\n",
    "    accuracy, precision, recall, f1 = calculate_metrics(y_test, y_pred)\n",
    "    accuracy_scores.append(accuracy)\n",
    "    precision_scores.append(precision)\n",
    "    recall_scores.append(recall)\n",
    "    f1_scores.append(f1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "e5a048c8-e023-413e-bb23-9febf7c0eb47",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Comparison of all ML models (Prolongation): \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>0.657789</td>\n",
       "      <td>0.676951</td>\n",
       "      <td>0.602585</td>\n",
       "      <td>0.637607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Decision Tree</td>\n",
       "      <td>0.575464</td>\n",
       "      <td>0.596674</td>\n",
       "      <td>0.463651</td>\n",
       "      <td>0.521818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>KNN</td>\n",
       "      <td>0.604520</td>\n",
       "      <td>0.601256</td>\n",
       "      <td>0.618740</td>\n",
       "      <td>0.609873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC</td>\n",
       "      <td>0.628733</td>\n",
       "      <td>0.623639</td>\n",
       "      <td>0.647819</td>\n",
       "      <td>0.635499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Naive Bayes</td>\n",
       "      <td>0.598063</td>\n",
       "      <td>0.600666</td>\n",
       "      <td>0.583199</td>\n",
       "      <td>0.591803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Logistic Regression</td>\n",
       "      <td>0.550444</td>\n",
       "      <td>0.548137</td>\n",
       "      <td>0.570275</td>\n",
       "      <td>0.558987</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Model  Accuracy  Precision    Recall  F1 Score\n",
       "0        Random Forest  0.657789   0.676951  0.602585  0.637607\n",
       "1        Decision Tree  0.575464   0.596674  0.463651  0.521818\n",
       "2                  KNN  0.604520   0.601256  0.618740  0.609873\n",
       "3                  SVC  0.628733   0.623639  0.647819  0.635499\n",
       "4          Naive Bayes  0.598063   0.600666  0.583199  0.591803\n",
       "5  Logistic Regression  0.550444   0.548137  0.570275  0.558987"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluation_df = pd.DataFrame({\n",
    "    'Model': models,\n",
    "    'Accuracy': accuracy_scores,\n",
    "    'Precision': precision_scores,\n",
    "    'Recall': recall_scores,\n",
    "    'F1 Score': f1_scores\n",
    "})\n",
    "print('Comparison of all ML models (Prolongation): ')\n",
    "evaluation_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "8764e477-c046-4838-bb2e-222a6409b0a3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Highest Accuracy score of 65.78% with Random Forest\n",
      "Highest Precision score of 67.70% with Random Forest\n",
      "Highest Recall score of 64.78% with SVC\n",
      "Highest F1 score of 63.76% with Random Forest\n"
     ]
    }
   ],
   "source": [
    "print(f'Highest Accuracy score of {np.max(accuracy_scores)*100:.2f}% with {models[np.argmax(accuracy_scores)]}')\n",
    "print(f'Highest Precision score of {np.max(precision_scores)*100:.2f}% with {models[np.argmax(precision_scores)]}')\n",
    "print(f'Highest Recall score of {np.max(recall_scores)*100:.2f}% with {models[np.argmax(recall_scores)]}')\n",
    "print(f'Highest F1 score of {np.max(f1_scores)*100:.2f}% with {models[np.argmax(f1_scores)]}')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}