Machine-Learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Goal\" data-toc-modified-id=\"Goal-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Goal</a></span></li><li><span><a href=\"#importing-the-libraries\" data-toc-modified-id=\"importing-the-libraries-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>importing the libraries</a></span></li><li><span><a href=\"#Plot-settings\" data-toc-modified-id=\"Plot-settings-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>Plot settings</a></span></li><li><span><a href=\"#Reading-the-data\" data-toc-modified-id=\"Reading-the-data-4\"><span class=\"toc-item-num\">4&nbsp;&nbsp;</span>Reading the data</a></span></li><li><span><a href=\"#Exploratory-data-analysis-(EDA)\" data-toc-modified-id=\"Exploratory-data-analysis-(EDA)-5\"><span class=\"toc-item-num\">5&nbsp;&nbsp;</span>Exploratory data analysis (EDA)</a></span><ul class=\"toc-item\"><li><span><a href=\"#Pairplot-of-the-all-the-features\" data-toc-modified-id=\"Pairplot-of-the-all-the-features-5.1\"><span class=\"toc-item-num\">5.1&nbsp;&nbsp;</span>Pairplot of the all the features</a></span></li><li><span><a href=\"#Pearson-correlation-and-feature-selection\" data-toc-modified-id=\"Pearson-correlation-and-feature-selection-5.2\"><span class=\"toc-item-num\">5.2&nbsp;&nbsp;</span>Pearson correlation and feature selection</a></span></li><li><span><a href=\"#Class-distribution-in-Diabetes-dataset\" data-toc-modified-id=\"Class-distribution-in-Diabetes-dataset-5.3\"><span class=\"toc-item-num\">5.3&nbsp;&nbsp;</span>Class distribution in Diabetes dataset</a></span></li></ul></li><li><span><a href=\"#Preparing-data-for-machine-learning\" data-toc-modified-id=\"Preparing-data-for-machine-learning-6\"><span class=\"toc-item-num\">6&nbsp;&nbsp;</span>Preparing data for machine learning</a></span><ul class=\"toc-item\"><li><span><a href=\"#Undersampling-|-TomekLinks\" data-toc-modified-id=\"Undersampling-|-TomekLinks-6.1\"><span class=\"toc-item-num\">6.1&nbsp;&nbsp;</span>Undersampling | TomekLinks</a></span></li></ul></li><li><span><a href=\"#Machine-learning-Models\" data-toc-modified-id=\"Machine-learning-Models-7\"><span class=\"toc-item-num\">7&nbsp;&nbsp;</span>Machine learning Models</a></span><ul class=\"toc-item\"><li><span><a href=\"#k-Nearest-Neighbors-(KNN)\" data-toc-modified-id=\"k-Nearest-Neighbors-(KNN)-7.1\"><span class=\"toc-item-num\">7.1&nbsp;&nbsp;</span>k-Nearest Neighbors (KNN)</a></span><ul class=\"toc-item\"><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.1.1\"><span class=\"toc-item-num\">7.1.1&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.1.2\"><span class=\"toc-item-num\">7.1.2&nbsp;&nbsp;</span>K-fold cross validation</a></span></li></ul></li><li><span><a href=\"#Logistic-Regression\" data-toc-modified-id=\"Logistic-Regression-7.2\"><span class=\"toc-item-num\">7.2&nbsp;&nbsp;</span>Logistic Regression</a></span><ul class=\"toc-item\"><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.2.1\"><span class=\"toc-item-num\">7.2.1&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.2.2\"><span class=\"toc-item-num\">7.2.2&nbsp;&nbsp;</span>K-fold cross validation</a></span></li></ul></li><li><span><a href=\"#Decision-Tree\" data-toc-modified-id=\"Decision-Tree-7.3\"><span class=\"toc-item-num\">7.3&nbsp;&nbsp;</span>Decision Tree</a></span><ul class=\"toc-item\"><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.3.1\"><span class=\"toc-item-num\">7.3.1&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.3.2\"><span class=\"toc-item-num\">7.3.2&nbsp;&nbsp;</span>K-fold cross validation</a></span></li><li><span><a href=\"#Feature-importance-in-Decision-trees\" data-toc-modified-id=\"Feature-importance-in-Decision-trees-7.3.3\"><span class=\"toc-item-num\">7.3.3&nbsp;&nbsp;</span>Feature importance in Decision trees</a></span></li></ul></li><li><span><a href=\"#Random-Forest\" data-toc-modified-id=\"Random-Forest-7.4\"><span class=\"toc-item-num\">7.4&nbsp;&nbsp;</span>Random Forest</a></span><ul class=\"toc-item\"><li><span><a href=\"#Feature-importance-in-Random-Forest\" data-toc-modified-id=\"Feature-importance-in-Random-Forest-7.4.1\"><span class=\"toc-item-num\">7.4.1&nbsp;&nbsp;</span>Feature importance in Random Forest</a></span></li><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.4.2\"><span class=\"toc-item-num\">7.4.2&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.4.3\"><span class=\"toc-item-num\">7.4.3&nbsp;&nbsp;</span>K-fold cross-validation</a></span></li></ul></li><li><span><a href=\"#Gradient-Boosting\" data-toc-modified-id=\"Gradient-Boosting-7.5\"><span class=\"toc-item-num\">7.5&nbsp;&nbsp;</span>Gradient Boosting</a></span><ul class=\"toc-item\"><li><span><a href=\"#Feature-importance-in-Gradient-Boosting\" data-toc-modified-id=\"Feature-importance-in-Gradient-Boosting-7.5.1\"><span class=\"toc-item-num\">7.5.1&nbsp;&nbsp;</span>Feature importance in Gradient Boosting</a></span></li><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.5.2\"><span class=\"toc-item-num\">7.5.2&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.5.3\"><span class=\"toc-item-num\">7.5.3&nbsp;&nbsp;</span>K-fold cross-validation</a></span></li></ul></li><li><span><a href=\"#Support-Vector-Machine-(SVM)\" data-toc-modified-id=\"Support-Vector-Machine-(SVM)-7.6\"><span class=\"toc-item-num\">7.6&nbsp;&nbsp;</span>Support Vector Machine (SVM)</a></span><ul class=\"toc-item\"><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.6.1\"><span class=\"toc-item-num\">7.6.1&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.6.2\"><span class=\"toc-item-num\">7.6.2&nbsp;&nbsp;</span>K-fold cross-validation</a></span></li></ul></li><li><span><a href=\"#Neural-Networks\" data-toc-modified-id=\"Neural-Networks-7.7\"><span class=\"toc-item-num\">7.7&nbsp;&nbsp;</span>Neural Networks</a></span><ul class=\"toc-item\"><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.7.1\"><span class=\"toc-item-num\">7.7.1&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.7.2\"><span class=\"toc-item-num\">7.7.2&nbsp;&nbsp;</span>K-fold cross-validation</a></span></li></ul></li><li><span><a href=\"#XGBoost\" data-toc-modified-id=\"XGBoost-7.8\"><span class=\"toc-item-num\">7.8&nbsp;&nbsp;</span>XGBoost</a></span><ul class=\"toc-item\"><li><span><a href=\"#Confusion-matrix\" data-toc-modified-id=\"Confusion-matrix-7.8.1\"><span class=\"toc-item-num\">7.8.1&nbsp;&nbsp;</span>Confusion matrix</a></span></li><li><span><a href=\"#K-fold-cross-validation\" data-toc-modified-id=\"K-fold-cross-validation-7.8.2\"><span class=\"toc-item-num\">7.8.2&nbsp;&nbsp;</span>K-fold cross-validation</a></span></li><li><span><a href=\"#Feature-importance-in-XGBoost\" data-toc-modified-id=\"Feature-importance-in-XGBoost-7.8.3\"><span class=\"toc-item-num\">7.8.3&nbsp;&nbsp;</span>Feature importance in XGBoost</a></span></li></ul></li></ul></li><li><span><a href=\"#Model-comparision\" data-toc-modified-id=\"Model-comparision-8\"><span class=\"toc-item-num\">8&nbsp;&nbsp;</span>Model comparision</a></span></li><li><span><a href=\"#Conclusion\" data-toc-modified-id=\"Conclusion-9\"><span class=\"toc-item-num\">9&nbsp;&nbsp;</span>Conclusion</a></span></li><li><span><a href=\"#In-Future\" data-toc-modified-id=\"In-Future-10\"><span class=\"toc-item-num\">10&nbsp;&nbsp;</span>In Future</a></span></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Goal  \n",
    "\n",
    "**Classification of effectiveness of the drug based on Lipinski molecular descriptors for ABL-kinase using Machine learning classfiers**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# importing the libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import sweetviz as sv\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import LabelBinarizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score, confusion_matrix\n",
    "\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['axes.labelsize'] = 20\n",
    "plt.rcParams['xtick.labelsize'] = 15\n",
    "plt.rcParams['ytick.labelsize'] = 15\n",
    "plt.rcParams['legend.fontsize'] = 23\n",
    "plt.rcParams['figure.titlesize'] = 26\n",
    "plt.rcParams['xtick.major.size'] = 10\n",
    "plt.rcParams['xtick.major.width'] = 1\n",
    "plt.rcParams['ytick.major.size'] = 10\n",
    "plt.rcParams['ytick.major.width'] = 1\n",
    "plt.rcParams['xtick.minor.width'] = 1\n",
    "plt.rcParams['ytick.minor.size'] = 5\n",
    "plt.rcParams['ytick.minor.width'] = 1\n",
    "plt.rcParams['xtick.minor.size'] = 5\n",
    "sns.set_style('ticks')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Reading the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "csv_file='abl_kinase-data.csv'\n",
    "df = pd.read_csv(csv_file,index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "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>index</th>\n",
       "      <th>molecule_chembl_id</th>\n",
       "      <th>canonical_smiles</th>\n",
       "      <th>standard_value</th>\n",
       "      <th>bioactivity_class</th>\n",
       "      <th>MW</th>\n",
       "      <th>LogP</th>\n",
       "      <th>NumHDonors</th>\n",
       "      <th>NumHAcceptors</th>\n",
       "      <th>norm_value</th>\n",
       "      <th>pIC50</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>CHEMBL281470</td>\n",
       "      <td>COc1ccc(-c2[nH]nc3c2C(=O)c2ccccc2-3)cc1</td>\n",
       "      <td>725000.0</td>\n",
       "      <td>inactive</td>\n",
       "      <td>276.295</td>\n",
       "      <td>3.2967</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>725000.0</td>\n",
       "      <td>3.139662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>CHEMBL13462</td>\n",
       "      <td>CC(=O)N[C@@H](Cc1ccc(OP(=O)(O)O)cc1)C(=O)N[C@@...</td>\n",
       "      <td>16000.0</td>\n",
       "      <td>inactive</td>\n",
       "      <td>569.592</td>\n",
       "      <td>2.3738</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>16000.0</td>\n",
       "      <td>4.795880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>CHEMBL414123</td>\n",
       "      <td>CC[C@H](C)[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H](...</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>inactive</td>\n",
       "      <td>1473.536</td>\n",
       "      <td>-1.0806</td>\n",
       "      <td>17.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>4.823909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>CHEMBL413629</td>\n",
       "      <td>CC[C@H](C)[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H](...</td>\n",
       "      <td>13500.0</td>\n",
       "      <td>inactive</td>\n",
       "      <td>1507.544</td>\n",
       "      <td>-0.3251</td>\n",
       "      <td>17.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>13500.0</td>\n",
       "      <td>4.869666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>CHEMBL301845</td>\n",
       "      <td>CC[C@H](NC(=O)[C@H](Cc1ccc(OP(=O)(O)O)cc1)NC(C...</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>inactive</td>\n",
       "      <td>525.583</td>\n",
       "      <td>2.9190</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>4.823909</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index molecule_chembl_id  \\\n",
       "0      0       CHEMBL281470   \n",
       "2      2        CHEMBL13462   \n",
       "3      3       CHEMBL414123   \n",
       "6      6       CHEMBL413629   \n",
       "7      7       CHEMBL301845   \n",
       "\n",
       "                                    canonical_smiles  standard_value  \\\n",
       "0            COc1ccc(-c2[nH]nc3c2C(=O)c2ccccc2-3)cc1        725000.0   \n",
       "2  CC(=O)N[C@@H](Cc1ccc(OP(=O)(O)O)cc1)C(=O)N[C@@...         16000.0   \n",
       "3  CC[C@H](C)[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H](...         15000.0   \n",
       "6  CC[C@H](C)[C@H](NC(=O)[C@@H]1CCCN1C(=O)[C@@H](...         13500.0   \n",
       "7  CC[C@H](NC(=O)[C@H](Cc1ccc(OP(=O)(O)O)cc1)NC(C...         15000.0   \n",
       "\n",
       "  bioactivity_class        MW    LogP  NumHDonors  NumHAcceptors  norm_value  \\\n",
       "0          inactive   276.295  3.2967         1.0            3.0    725000.0   \n",
       "2          inactive   569.592  2.3738         5.0            6.0     16000.0   \n",
       "3          inactive  1473.536 -1.0806        17.0           19.0     15000.0   \n",
       "6          inactive  1507.544 -0.3251        17.0           18.0     13500.0   \n",
       "7          inactive   525.583  2.9190         4.0            5.0     15000.0   \n",
       "\n",
       "      pIC50  \n",
       "0  3.139662  \n",
       "2  4.795880  \n",
       "3  4.823909  \n",
       "6  4.869666  \n",
       "7  4.823909  "
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exploratory data analysis (EDA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dimension of diabetes data: (2159, 11)\n"
     ]
    }
   ],
   "source": [
    "print(\"dimension of diabetes data: {}\".format(df.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 2159 entries, 0 to 2436\n",
      "Data columns (total 11 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   index               2159 non-null   int64  \n",
      " 1   molecule_chembl_id  2159 non-null   object \n",
      " 2   canonical_smiles    2159 non-null   object \n",
      " 3   standard_value      2159 non-null   float64\n",
      " 4   bioactivity_class   2159 non-null   object \n",
      " 5   MW                  2159 non-null   float64\n",
      " 6   LogP                2159 non-null   float64\n",
      " 7   NumHDonors          2159 non-null   float64\n",
      " 8   NumHAcceptors       2159 non-null   float64\n",
      " 9   norm_value          2159 non-null   float64\n",
      " 10  pIC50               2159 non-null   float64\n",
      "dtypes: float64(7), int64(1), object(3)\n",
      "memory usage: 202.4+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " * Dataset is small and well labeled. There are no null values present.\n",
    " * very suitable to supervised machine learning formulation.\n",
    " * This is a binary classification problem, where we have 2 classes in the target **(y)** (i.e.`'bioactivity_class'`) and the Lipinski Molecular descriptors (i.e. `MW`,`LogP`,`NumHDonors` and `NumHAcceptors` can be used as the feature (**X**)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "selection = ['MW','LogP','NumHDonors','NumHAcceptors','bioactivity_class']\n",
    "df = df[selection]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pairplot of the all the features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 880.925x720 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "selection = ['bioactivity_class','MW','LogP','NumHDonors','NumHAcceptors']\n",
    "sns.pairplot(df[selection], hue='bioactivity_class')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pearson correlation and feature selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,7))\n",
    "corr_mat = df.iloc[:,:-1].corr()\n",
    "sns.heatmap(corr_mat, fmt='0.2f', annot=True, lw=2, cbar_kws={'label':'Pearson Correlation (r)'})\n",
    "plt.xticks(size=15,rotation=90)\n",
    "plt.yticks(size=15,rotation=0)\n",
    "plt.tight_layout()\n",
    "plt.savefig('Correlation.png',dpi=300);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* All the features hold very little correlation amongst each other. Only **NumHAcceptors** and **Molecular weight** shows a significant  strong correlation. Therefore, keeping the all the features. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Class distribution in Diabetes dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,5))\n",
    "sns.countplot(df['bioactivity_class'],label='Count')\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.ylabel('Count',size=15)\n",
    "plt.xlabel('Class',size=15)\n",
    "sns.despine(top=True)\n",
    "plt.title('Class distribution',size=26, weight='bold')\n",
    "plt.tight_layout()\n",
    "plt.savefig('Class-distribution.png');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preparing data for machine learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['bioactivity_class'] = df['bioactivity_class'].map({'inactive':0, 'active':1})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.iloc[:,:-1].values   ##features selection\n",
    "y = df.iloc[:,-1].values    ## target selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X,y, stratify=y, random_state=27)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* It is importannt to use stratify inside **train_test_split**. It keeps the same distribution of target same  within train and testing datasets. if variable `y` is a binary categorical variable with values 0 and 1 and there are 25% of zeros and 75% of ones, `stratify=y` will make sure that random split has 40% of 0's and 60% of 1's"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Undersampling | TomekLinks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [],
   "source": [
    "rs = SMOTE(sampling_strategy='auto')\n",
    "X_train,y_train = rs.fit_resample(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x128b102d0>"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,5))\n",
    "sns.countplot(y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Machine learning Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k-Nearest Neighbors (KNN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "working on neighbors 2\n",
      "working on neighbors 3\n",
      "working on neighbors 4\n",
      "working on neighbors 5\n",
      "working on neighbors 6\n",
      "working on neighbors 7\n",
      "working on neighbors 8\n",
      "working on neighbors 9\n",
      "working on neighbors 10\n",
      "working on neighbors 11\n",
      "working on neighbors 12\n",
      "working on neighbors 13\n",
      "working on neighbors 14\n",
      "working on neighbors 15\n",
      "working on neighbors 16\n",
      "working on neighbors 17\n",
      "working on neighbors 18\n",
      "working on neighbors 19\n"
     ]
    }
   ],
   "source": [
    "training_accuracy = [] \n",
    "test_accuracy = []\n",
    "training_f1 = []\n",
    "test_f1 = []\n",
    "\n",
    "neighbors_settings = range(2,20)\n",
    "\n",
    "for n_neighbors in neighbors_settings:\n",
    "    print(f'working on neighbors {n_neighbors}')\n",
    "    knn = KNeighborsClassifier(n_neighbors=n_neighbors)\n",
    "    knn.fit(X_train, y_train)\n",
    "    \n",
    "    y_train_pred=knn.predict(X_train)\n",
    "    y_pred=knn.predict(X_test)\n",
    "    \n",
    "    training_accuracy.append(accuracy_score(y_train,y_train_pred))\n",
    "    test_accuracy.append(accuracy_score(y_test, y_pred))\n",
    "    \n",
    "    training_f1.append(f1_score(y_train,y_train_pred))\n",
    "    test_f1.append(f1_score(y_test, y_pred))\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(14,10))\n",
    "fig.subplots_adjust(hspace=0.5, wspace=0.3)\n",
    "\n",
    "fig.add_subplot(2,2,1)\n",
    "plt.plot(neighbors_settings, training_accuracy, label='training accuracy')\n",
    "plt.plot(neighbors_settings, test_accuracy, label='test accuracy')\n",
    "plt.ylabel('Accuracy',size=20)\n",
    "plt.xticks(range(2,20,2),size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.xlabel('n_neighbors',size=20)\n",
    "plt.title('Accuracy Score',size=20, weight='bold')\n",
    "plt.legend([],frameon=False, prop={'size':20})\n",
    "\n",
    "fig.add_subplot(2,2,2)\n",
    "plt.plot(neighbors_settings, training_f1)\n",
    "plt.plot(neighbors_settings, test_f1)\n",
    "plt.xticks(range(2,20,2),size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.ylabel('F1 Score',size=20)\n",
    "plt.xlabel('n_neighbors',size=20)\n",
    "plt.title('F1-Score',size=20,weight='bold')\n",
    "plt.legend(['Training','Testing'],frameon=False, bbox_to_anchor=(0.2,-0.2), ncol=2, prop={'size':20});\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`Accuracy score`, `F1-score` suggest that `n_neighbors=5` is the optimum choice **(by looking a testing accuracies)**. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of K-NN classifier on test set: 0.91\n",
      "Accuracy of K-NN classifier on test set: 0.77\n",
      "F1-score of K-NN classifier on test set: 0.85\n"
     ]
    }
   ],
   "source": [
    "knn = KNeighborsClassifier(n_neighbors=5).fit(X_train, y_train)\n",
    "y_pred = knn.predict(X_test)\n",
    "\n",
    "print('Accuracy of K-NN classifier on test set: {:.2f}'.format(knn.score(X_train, y_train)))\n",
    "print('Accuracy of K-NN classifier on test set: {:.2f}'.format(accuracy_score(y_test, y_pred)))\n",
    "print('F1-score of K-NN classifier on test set: {:.2f}'.format(f1_score(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalized_confusion_matrix(y_test, conf_mat, model):\n",
    "    _ , counts = np.unique(y_test,return_counts=True)\n",
    "    conf_mat = (conf_mat.T/counts).T\n",
    "    plt.figure(figsize=(6,5))\n",
    "    ax=sns.heatmap(conf_mat,fmt='.2f',annot=True,annot_kws={'size':20},lw=2, cbar=True, cbar_kws={'label':'% Class accuracy'})\n",
    "    plt.title(f'Confusion Matrix ({model})',size=22)\n",
    "    plt.xticks(size=20)\n",
    "    plt.yticks(size=20)\n",
    "    ax.figure.axes[-1].yaxis.label.set_size(20) ##colorbar label\n",
    "    cax = plt.gcf().axes[-1]  ##colorbar ticks\n",
    "    cax.tick_params(labelsize=20) ## colorbar ticks\n",
    "    plt.savefig(f'confusion-matrix-{model}.png',dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat, 'kNN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_knn_accuracy = cross_val_score(knn, X, y, cv=10) ##10-fold cross validation\n",
    "k_fold_knn_f1 = cross_val_score(knn, X, y, cv=10,scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation :0.77 +/- 0.08\n",
      "Average F1-score after 10 fold cross validation :0.75 +/- 0.05\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation :{k_fold_knn_accuracy.mean().round(2)} +/- {k_fold_knn_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation :{k_fold_knn_f1.mean().round(2)} +/- {k_fold_knn_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C : 1 Training set accuracy: 0.639\n",
      "C : 1 Test set accuracy: 0.663\n",
      "C : 1 Training set F1-score: 0.645\n",
      "C : 1 Test set F1-score: 0.766\n",
      "\n",
      "\n",
      "C : 0.01 Training set accuracy: 0.641\n",
      "C : 0.01 Test set accuracy: 0.669\n",
      "C : 0.01 Training set F1-score: 0.649\n",
      "C : 0.01 Test set F1-score: 0.771\n",
      "\n",
      "\n",
      "C : 100 Training set accuracy: 0.639\n",
      "C : 100 Test set accuracy: 0.663\n",
      "C : 100 Training set F1-score: 0.645\n",
      "C : 100 Test set F1-score: 0.766\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1269ed910>"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,6))\n",
    "Clist=[1,0.01,100]\n",
    "\n",
    "for C in Clist : \n",
    "\n",
    "    logreg = LogisticRegression(C=C,solver='newton-cg').fit(X_train, y_train) #keeping C=1 a\n",
    "    y_train_pred = logreg.predict(X_train)\n",
    "    y_pred = logreg.predict(X_test)\n",
    "\n",
    "    print('C : {} Training set accuracy: {:.3f}'.format(C,accuracy_score(y_train, y_train_pred)))\n",
    "    print('C : {} Test set accuracy: {:.3f}'.format(C,accuracy_score(y_test, y_pred)))\n",
    "\n",
    "    print('C : {} Training set F1-score: {:.3f}'.format(C,f1_score(y_train, y_train_pred)))\n",
    "    print('C : {} Test set F1-score: {:.3f}'.format(C, f1_score(y_test, y_pred)))\n",
    "    print('\\n')\n",
    "    \n",
    "    data_features = [x for i,x in enumerate(df.columns[:-1]) if i!=8]\n",
    "    plt.plot(logreg.coef_.T, marker='o', label=f\"C={C}\")\n",
    "\n",
    "plt.xticks(range(df.shape[1]), data_features, rotation=90)\n",
    "plt.hlines(0, 0, df.shape[1])\n",
    "plt.ylim(-5, 5)\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.xlabel(\"Feature\",size=20)\n",
    "plt.ylabel(\"Coefficient magnitude\",size=15)\n",
    "plt.legend(frameon=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The default value of C=1 provides with 81% accuracy on training and 82% accuracy on test set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Stronger regularization (`C=0.01`) pushes coefficients more and more toward zero. Inspecting the plot more closely suggests that feature `logP`, for `C=100`, `C=1` and `C=0.001`, the coefficient is **positive**. This indicates that high `LogP` feature is related to a sample being`diabetes`, regardless which model we look at."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [],
   "source": [
    "logreg = LogisticRegression(C=1,solver='newton-cg').fit(X_train,y_train)\n",
    "y_pred=logreg.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'Logistic Regression')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_logreg_accuracy = cross_val_score(logreg, X, y, cv=10) ##10-fold cross validation\n",
    "k_fold_logreg_f1 = cross_val_score(logreg, X, y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation :0.79 +/- 0.09\n",
      "Average F1-score after 10 fold cross validation :0.72 +/- 0.05\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation :{k_fold_logreg_accuracy.mean().round(2)} +/- {k_fold_logreg_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation :{k_fold_logreg_f1.mean().round(2)} +/- {k_fold_logreg_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_depth=range(1,20)\n",
    "training_accuracy = [] \n",
    "test_accuracy = []\n",
    "training_f1 = []\n",
    "test_f1 = []\n",
    "\n",
    "for depth in max_depth : \n",
    "    tree = DecisionTreeClassifier(random_state=0, max_depth=depth, min_samples_leaf=1).fit(X_train, y_train)\n",
    "    y_train_pred = tree.predict(X_train)\n",
    "    y_pred = tree.predict(X_test)\n",
    "    \n",
    "    training_accuracy.append(accuracy_score(y_train,y_train_pred))\n",
    "    test_accuracy.append(accuracy_score(y_test, y_pred))\n",
    "    \n",
    "    training_f1.append(f1_score(y_train,y_train_pred))\n",
    "    test_f1.append(f1_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(14,10))\n",
    "fig.subplots_adjust(hspace=0.5, wspace=0.3)\n",
    "\n",
    "fig.add_subplot(2,2,1)\n",
    "plt.plot(max_depth, training_accuracy, label='training accuracy')\n",
    "plt.plot(max_depth, test_accuracy, label='test accuracy')\n",
    "plt.ylabel('Accuracy',size=20)\n",
    "plt.xlabel('Max depth',size=20)\n",
    "plt.xticks(range(1,20,2),size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.title('Accuracy Score',size=20, weight='bold')\n",
    "plt.legend([],frameon=False)\n",
    "\n",
    "fig.add_subplot(2,2,2)\n",
    "plt.plot(max_depth, training_f1)\n",
    "plt.plot(max_depth, test_f1)\n",
    "plt.ylabel('F1 Score',size=20)\n",
    "plt.xlabel('Max depth',size=20)\n",
    "plt.xticks(range(1,20,2),size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.title('F1-Score',size=20,weight='bold')\n",
    "plt.legend(['Training','Testing'],frameon=False, bbox_to_anchor=(0.3,-0.2), prop = {'size':20}, ncol=2);\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The training accuracy on the after training set is 100%, while the test set accuracy is much worse. This means tree is  overfitting and not generalizing well to new data. We set `max_depth=9`, limiting the depth of the tree decreases overfitting. This leads to a lower accuracy on the training set, but an improvement on the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on test: 0.817\n",
      "F1-score on test set: 0.885\n"
     ]
    }
   ],
   "source": [
    "tree = DecisionTreeClassifier(max_depth=9, min_samples_leaf=1, random_state=0).fit(X_train, y_train)\n",
    "y_pred=tree.predict(X_test)\n",
    "\n",
    "print(\"Accuracy on test: {:.3f}\".format(accuracy_score(y_pred, y_test)))\n",
    "print(\"F1-score on test set: {:.3f}\".format(f1_score(y_pred, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'Decision Tree')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_tree_accuracy = cross_val_score(tree, X, y, cv=10) ##10-fold cross validation\n",
    "k_fold_tree_f1 = cross_val_score(tree, X, y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation :0.77 +/- 0.1\n",
      "Average F1-score after 10 fold cross validation :0.77 +/- 0.08\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation :{k_fold_tree_accuracy.mean().round(2)} +/- {k_fold_tree_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation :{k_fold_tree_f1.mean().round(2)} +/- {k_fold_tree_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importance in Decision trees\n",
    "Feature importance rates how important each feature is for the decision a tree makes. It is a number between 0 and 1\n",
    "\n",
    "     0 means “not used at all” \n",
    "     1 means “perfectly predicts the target”"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_feature_importances(model, figure):\n",
    "    n_features = 4\n",
    "    plt.figure(figsize=(10,6))\n",
    "    plt.barh(range(n_features), model.feature_importances_, align='center')\n",
    "    plt.yticks(np.arange(n_features), data_features,size=15)\n",
    "    plt.xticks(size=15)\n",
    "    plt.xlabel('Feature importance',size=20,)\n",
    "    #plt.ylabel('Feature',size=20)\n",
    "    plt.ylim(-1, n_features)\n",
    "    sns.despine(top=True)\n",
    "    plt.title(f'{figure}',size=20)\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(f'feature-image{figure}.png',dpi=300)\n",
    "\n",
    "plot_feature_importances(tree,'Decision Tree')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Feature `MW` is the most important feature."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on test: 0.856\n",
      "F1-score on test set: 0.911\n"
     ]
    }
   ],
   "source": [
    "rf = RandomForestClassifier(n_estimators=100, random_state=0).fit(X_train, y_train)\n",
    "y_pred=rf.predict(X_test)\n",
    "\n",
    "print(\"Accuracy on test: {:.3f}\".format(accuracy_score(y_pred, y_test)))\n",
    "print(\"F1-score on test set: {:.3f}\".format(f1_score(y_pred, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " The default parameters of the random forest work well. This conclusion is derived after trying various values of  `max_depth` and `n_estimators`. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importance in Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_feature_importances(rf, 'Random Forest')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Similarly to the single decision tree, the random forest also gives a lot of importance to the `LogP` feature and it also chooses `MW` to be the 2nd most informative feature. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'Random Forest')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_rf_accuracy = cross_val_score(rf, X, y, cv=10) ##10-fold cross validation\n",
    "k_fold_rf_f1 = cross_val_score(rf, X, y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation :0.77 +/- 0.1\n",
      "Average F1-score after 10 fold cross validation :0.77 +/- 0.08\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation :{k_fold_tree_accuracy.mean().round(2)} +/- {k_fold_tree_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation :{k_fold_tree_f1.mean().round(2)} +/- {k_fold_tree_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on train: 0.916\n",
      "Accuracy on test: 0.870\n",
      "F1-score on test set: 0.923\n"
     ]
    }
   ],
   "source": [
    "gb = GradientBoostingClassifier(random_state=0).fit(X_train, y_train)\n",
    "y_pred=gb.predict(X_test)\n",
    "\n",
    "print(\"Accuracy on train: {:.3f}\".format(gb.score(X_train, y_train)))\n",
    "print(\"Accuracy on test: {:.3f}\".format(accuracy_score(y_pred, y_test)))\n",
    "print(\"F1-score on test set: {:.3f}\".format(f1_score(y_pred, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Default values result in **overfitting**. Therefore, `max_depth = 1` and `learning_rate = 0.01` is chosen which reduces the gap between training and test accuracy. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importance in Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_feature_importances(gb, 'Gradient Boosting')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Gradient Boosting gives the most importance to `NumHDonors` and again 2nd best is `MW`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'Gradient Boosting')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_gb_accuracy = cross_val_score(gb, X, y, cv=10) ##10-fold cross validation\n",
    "k_fold_gb_f1 = cross_val_score(gb, X, y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation :0.81 +/- 0.08\n",
      "Average F1-score after 10 fold cross validation :0.79 +/- 0.06\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation :{k_fold_gb_accuracy.mean().round(2)} +/- {k_fold_gb_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation :{k_fold_gb_f1.mean().round(2)} +/- {k_fold_gb_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Support Vector Machine (SVM)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on train: 0.734\n",
      "Accuracy on test: 0.739\n",
      "F1-score on test set: 0.822\n"
     ]
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train)\n",
    "X_test_scaled = scaler.fit_transform(X_test)\n",
    "\n",
    "#SVM requiest feature scaling\n",
    "svc = SVC().fit(X_train_scaled, y_train)\n",
    "y_pred=svc.predict(X_test_scaled)\n",
    "\n",
    "print(\"Accuracy on train: {:.3f}\".format(svc.score(X_train_scaled, y_train)))\n",
    "print(\"Accuracy on test: {:.3f}\".format(accuracy_score(y_pred, y_test)))\n",
    "print(\"F1-score on test set: {:.3f}\".format(f1_score(y_pred, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'SVM')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_svm_accuracy = cross_val_score(svc, scaler.fit_transform(X), y, cv=10) ##10-fold cross validation\n",
    "k_fold_svm_f1 = cross_val_score(svc, scaler.fit_transform(X), y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation : 0.83 +/- 0.05\n",
      "Average F1-score after 10 fold cross validation : 0.78 +/- 0.05\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation : {k_fold_svm_accuracy.mean().round(2)} +/- {k_fold_svm_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation : {k_fold_svm_f1.mean().round(2)} +/- {k_fold_svm_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Neural Networks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on train: 0.847\n",
      "Accuracy on test: 0.765\n",
      "F1-score on test set: 0.841\n"
     ]
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train)\n",
    "X_test_scaled = scaler.fit_transform(X_test)\n",
    "\n",
    "mlp = MLPClassifier(random_state=0,max_iter=2000).fit(X_train_scaled, y_train)\n",
    "y_pred=mlp.predict(X_test_scaled)\n",
    "\n",
    "print(\"Accuracy on train: {:.3f}\".format(mlp.score(X_train_scaled, y_train)))\n",
    "print(\"Accuracy on test: {:.3f}\".format(accuracy_score(y_pred, y_test)))\n",
    "print(\"F1-score on test set: {:.3f}\".format(f1_score(y_pred, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'Neural Networks')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_mlp_accuracy = cross_val_score(mlp, scaler.fit_transform(X), y, cv=10) ##10-fold cross validation\n",
    "k_fold_mlp_f1 = cross_val_score(mlp, scaler.fit_transform(X), y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation : 0.78 +/- 0.09\n",
      "Average F1-score after 10 fold cross validation : 0.77 +/- 0.07\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation : {k_fold_mlp_accuracy.mean().round(2)} +/- {k_fold_mlp_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation : {k_fold_mlp_f1.mean().round(2)} +/- {k_fold_mlp_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on train: 0.992\n",
      "Accuracy on test: 0.887\n",
      "F1-score on test set: 0.931\n"
     ]
    }
   ],
   "source": [
    "xgb = XGBClassifier().fit(X_train,y_train)\n",
    "\n",
    "y_pred=xgb.predict(X_test)\n",
    "\n",
    "print(\"Accuracy on train: {:.3f}\".format(xgb.score(X_train, y_train)))\n",
    "print(\"Accuracy on test: {:.3f}\".format(xgb.score(X_test, y_test)))\n",
    "print(\"F1-score on test set: {:.3f}\".format(f1_score(y_pred, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(y_test,y_pred)\n",
    "normalized_confusion_matrix(y_test,conf_mat,'XGBoost')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-fold cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_fold_xgb_accuracy = cross_val_score(xgb, X, y, cv=10) ##10-fold cross validation\n",
    "k_fold_xgb_f1 = cross_val_score(xgb, X, y, cv=10, scoring='f1_weighted') ##10-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average accuracy after 10 fold cross validation : 0.79 +/- 0.09\n",
      "Average F1-score after 10 fold cross validation : 0.79 +/- 0.07\n"
     ]
    }
   ],
   "source": [
    "print(f'Average accuracy after 10 fold cross validation : {k_fold_xgb_accuracy.mean().round(2)} +/- {k_fold_xgb_accuracy.std().round(2)}')\n",
    "print(f'Average F1-score after 10 fold cross validation : {k_fold_xgb_f1.mean().round(2)} +/- {k_fold_xgb_f1.std().round(2)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importance in XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_feature_importances(xgb, 'XGBoost')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model comparision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_accuracies = [k_fold_knn_accuracy.mean().round(2),\n",
    "                   k_fold_logreg_accuracy.mean().round(2),\n",
    "                   k_fold_tree_accuracy.mean().round(2),\n",
    "                   k_fold_rf_accuracy.mean().round(2),\n",
    "                   k_fold_gb_accuracy.mean().round(2),\n",
    "                   k_fold_svm_accuracy.mean().round(2),\n",
    "                   k_fold_mlp_accuracy.mean().round(2),\n",
    "                   k_fold_xgb_accuracy.mean().round(2)]\n",
    "\n",
    "all_accuracies_errors = [k_fold_knn_accuracy.std().round(2),\n",
    "                   k_fold_logreg_accuracy.std().round(2),\n",
    "                   k_fold_tree_accuracy.std().round(2),\n",
    "                   k_fold_rf_accuracy.std().round(2),\n",
    "                   k_fold_gb_accuracy.std().round(2),\n",
    "                   k_fold_svm_accuracy.std().round(2),\n",
    "                   k_fold_mlp_accuracy.std().round(2),\n",
    "                   k_fold_xgb_accuracy.std().round(2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_f1 = [k_fold_knn_f1.mean().round(2),\n",
    "                   k_fold_logreg_f1.mean().round(2),\n",
    "                   k_fold_tree_f1.mean().round(2),\n",
    "                   k_fold_rf_f1.mean().round(2),\n",
    "                   k_fold_gb_f1.mean().round(2),\n",
    "                   k_fold_svm_f1.mean().round(2),\n",
    "                   k_fold_mlp_f1.mean().round(2),\n",
    "                   k_fold_xgb_f1.mean().round(2)]\n",
    "\n",
    "all_f1_errors = [k_fold_knn_f1.std().round(2),\n",
    "                   k_fold_logreg_f1.std().round(2),\n",
    "                   k_fold_tree_f1.std().round(2),\n",
    "                   k_fold_rf_f1.std().round(2),\n",
    "                   k_fold_gb_f1.std().round(2),\n",
    "                   k_fold_svm_f1.std().round(2),\n",
    "                   k_fold_mlp_f1.std().round(2),\n",
    "                   k_fold_xgb_f1.std().round(2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {},
   "outputs": [],
   "source": [
    "models=['kNN','Logistic Regression','Decision Tree','Random Forest','Gradient Boosting','SVM','Neural Networks', 'XGBoost']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_data = pd.DataFrame([all_accuracies,all_accuracies_errors,all_f1,all_f1_errors],columns=models, index = ['Accuracy','STD_acc','F1-macro','STD_f1']).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col0 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col1 {\n",
       "            background-color:  #f2cbb7;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col2 {\n",
       "            background-color:  #c9d7f0;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col0 {\n",
       "            background-color:  #aac7fd;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col1 {\n",
       "            background-color:  #ee8468;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col2 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col0 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col1 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col2 {\n",
       "            background-color:  #f7a889;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col3 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col0 {\n",
       "            background-color:  #aac7fd;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col1 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col2 {\n",
       "            background-color:  #e26952;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col3 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col0 {\n",
       "            background-color:  #f7b89c;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col1 {\n",
       "            background-color:  #f2cbb7;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col2 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col3 {\n",
       "            background-color:  #aac7fd;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col0 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col1 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col2 {\n",
       "            background-color:  #e26952;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col0 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col1 {\n",
       "            background-color:  #ee8468;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col2 {\n",
       "            background-color:  #f7a889;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col3 {\n",
       "            background-color:  #f7b89c;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col0 {\n",
       "            background-color:  #aac7fd;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col1 {\n",
       "            background-color:  #ee8468;\n",
       "            color:  #000000;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col2 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col3 {\n",
       "            background-color:  #f7b89c;\n",
       "            color:  #000000;\n",
       "        }</style><table id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89b\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >Accuracy</th>        <th class=\"col_heading level0 col1\" >STD_acc</th>        <th class=\"col_heading level0 col2\" >F1-macro</th>        <th class=\"col_heading level0 col3\" >STD_f1</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row0\" class=\"row_heading level0 row0\" >kNN</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col0\" class=\"data row0 col0\" >0.770000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col1\" class=\"data row0 col1\" >0.080000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col2\" class=\"data row0 col2\" >0.750000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow0_col3\" class=\"data row0 col3\" >0.050000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row1\" class=\"row_heading level0 row1\" >Logistic Regression</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col0\" class=\"data row1 col0\" >0.790000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col1\" class=\"data row1 col1\" >0.090000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col2\" class=\"data row1 col2\" >0.720000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow1_col3\" class=\"data row1 col3\" >0.050000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row2\" class=\"row_heading level0 row2\" >Decision Tree</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col0\" class=\"data row2 col0\" >0.770000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col1\" class=\"data row2 col1\" >0.100000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col2\" class=\"data row2 col2\" >0.770000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow2_col3\" class=\"data row2 col3\" >0.080000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row3\" class=\"row_heading level0 row3\" >Random Forest</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col0\" class=\"data row3 col0\" >0.790000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col1\" class=\"data row3 col1\" >0.100000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col2\" class=\"data row3 col2\" >0.780000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow3_col3\" class=\"data row3 col3\" >0.080000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row4\" class=\"row_heading level0 row4\" >Gradient Boosting</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col0\" class=\"data row4 col0\" >0.810000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col1\" class=\"data row4 col1\" >0.080000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col2\" class=\"data row4 col2\" >0.790000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow4_col3\" class=\"data row4 col3\" >0.060000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row5\" class=\"row_heading level0 row5\" >SVM</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col0\" class=\"data row5 col0\" >0.830000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col1\" class=\"data row5 col1\" >0.050000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col2\" class=\"data row5 col2\" >0.780000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow5_col3\" class=\"data row5 col3\" >0.050000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row6\" class=\"row_heading level0 row6\" >Neural Networks</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col0\" class=\"data row6 col0\" >0.780000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col1\" class=\"data row6 col1\" >0.090000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col2\" class=\"data row6 col2\" >0.770000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow6_col3\" class=\"data row6 col3\" >0.070000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89blevel0_row7\" class=\"row_heading level0 row7\" >XGBoost</th>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col0\" class=\"data row7 col0\" >0.790000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col1\" class=\"data row7 col1\" >0.090000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col2\" class=\"data row7 col2\" >0.790000</td>\n",
       "                        <td id=\"T_28719d0a_ae8d_11ea_b118_6c96cfdbc89brow7_col3\" class=\"data row7 col3\" >0.070000</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1254b9cd0>"
      ]
     },
     "execution_count": 273,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_data.style.background_gradient(cmap='coolwarm',axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "color=sns.color_palette('Set2')\n",
    "model_data[['Accuracy','F1-macro']].plot.barh(figsize=(10,7),color=color)\n",
    "plt.legend(frameon=False,bbox_to_anchor=(1,0.8), prop={'size':20})\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.xlim([0.72,0.85]);\n",
    "plt.title('Model performances after K-fold cross-validation',size=20)\n",
    "sns.despine(top=True)\n",
    "plt.tight_layout()\n",
    "plt.savefig('model-comparision.png',dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "color=sns.color_palette('hls')\n",
    "plt.figure(figsize=(10,7))\n",
    "drug_accuracy = [0.62, 0.59, 0.59, 0.63, 0.52, 0.73, 0.75, 0.64]\n",
    "plt.barh(models,drug_accuracy,alpha=0.5, color=color[5])\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.xlim([0.4,0.8])\n",
    "plt.title('Accuracy in Drug non-Effectiveness',size=20)\n",
    "sns.despine(top=True)\n",
    "plt.tight_layout()\n",
    "plt.savefig('drug-noneffectiveness.png',dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,7))\n",
    "drug_accuracy = [0.80, 0.68, 0.87, 0.91, 0.95, 0.74, 0.77, 0.94]\n",
    "plt.barh(models,drug_accuracy,alpha=0.5)\n",
    "plt.xlim([0.4,0.95])\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.title('Accuracy in Drug Effectiveness',size=20)\n",
    "sns.despine(top=True)\n",
    "plt.tight_layout()\n",
    "plt.savefig('drug-effectiveness.png',dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* All model seems to provide the decent performance based on **10-fold cross validation** of the dataset. `Gradient boosting`seems to providing the best performance.\n",
    "\n",
    "* **Neural Network** achieves a highest score in predicting both classes.\n",
    "* **Feature selection** suggests the `NumHDonors` and `MW` are the most crucial factor for the successful prediction of bioactivity of the drug. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  In Future"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* More hyperparameter tuning.\n",
    "* More classifiers such as `catboost` to improve the performance.\n",
    "* Generating a Tableau Dashboard"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.4"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "288px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}