4a_Random_Forest.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "414e4df0-f5d2-4198-bb5f-ba6a9417d3d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from joblib import dump, load\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score,classification_report,confusion_matrix, ConfusionMatrixDisplay\n",
    "from sklearn.model_selection import RandomizedSearchCV"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d12450d4-aedd-4744-a44a-f84046e2a885",
   "metadata": {},
   "source": [
    "# Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "86d17afa-a7b8-4a59-adb3-89673f70bec3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.read_csv(\"X_train_stemm_cleaned.csv\")\n",
    "df_valid = pd.read_csv(\"X_valid_stemm_cleaned.csv\")\n",
    "df_test = pd.read_csv(\"X_test_stemm_cleaned.csv\")\n",
    "\n",
    "X_train_clean = df_train['tweet_stemming']\n",
    "X_valid_clean = df_valid['tweet_stemming']\n",
    "X_test_clean = df_test['tweet_stemming']\n",
    "\n",
    "y_train = pd.read_csv(\"y_train.csv\")\n",
    "y_valid = pd.read_csv(\"y_valid.csv\")\n",
    "y_test = pd.read_csv(\"y_test.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e70fdfbd-cb3f-446d-b044-df3d6f26cf36",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load model TF-IDF\n",
    "tfidf_vectorizer = load('tfidf_vectorizer.joblib')\n",
    "\n",
    "# transform set data dari load model tfidf yang telah disimpan\n",
    "X_train_tfidf = tfidf_vectorizer.transform(X_train_clean)\n",
    "X_valid_tfidf = tfidf_vectorizer.transform(X_valid_clean)\n",
    "X_test_tfidf = tfidf_vectorizer.transform(X_test_clean)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6d0209c-c894-4332-a726-3d8d2a0d5117",
   "metadata": {},
   "source": [
    "# Model Intialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "fa869733-5ae5-4f50-a15f-8dbeeb298ed1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# insialisasi model random forest untuk klasifikasi\n",
    "random_forest = RandomForestClassifier(n_estimators = 100,\n",
    "                                       criterion = \"gini\",\n",
    "                                       max_features = \"sqrt\",\n",
    "                                       max_depth=6,\n",
    "                                       min_samples_leaf=8,\n",
    "                                       random_state = 123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "c724bdbf-b0ae-4e0d-8cd0-72f809daa671",
   "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=6, min_samples_leaf=8, random_state=123)</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-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(max_depth=6, min_samples_leaf=8, random_state=123)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestClassifier(max_depth=6, min_samples_leaf=8, random_state=123)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit random forest model klasifikasi\n",
    "random_forest.fit(X_train_tfidf, np.ravel(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f794b562-684a-46c1-af23-61de6908cc54",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6543307086614173"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# score accuracy pada data train\n",
    "random_forest.score(X_train_tfidf, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c6b4d976-c5a7-49ce-bb91-ec9931dab0dc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5551470588235294"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#score accuracy pada data test\n",
    "random_forest.score(X_test_tfidf, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "86da7eaa-bd70-461b-975d-d98c3c31d460",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_rf = random_forest.predict(X_test_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "639bf3cd-5166-411c-b43a-ed83800a5ece",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "     negatif       0.59      0.57      0.58        89\n",
      "      netral       0.54      0.57      0.55        91\n",
      "     positif       0.55      0.52      0.53        92\n",
      "\n",
      "    accuracy                           0.56       272\n",
      "   macro avg       0.56      0.56      0.56       272\n",
      "weighted avg       0.56      0.56      0.56       272\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_rf))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d649317b-d4df-4464-a922-1fb7f844a5b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "unique_labels = np.unique(y_test)\n",
    "ConfusionMatrixDisplay.from_predictions(y_test, y_pred_rf, display_labels=unique_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f04468aa-5cab-468c-8880-f8387bceec30",
   "metadata": {},
   "source": [
    "# Hyperparameter Tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "2a5401c8-84f1-440d-a7dd-3df89b557e97",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters found:  {'n_estimators': 200, 'min_samples_split': 5, 'min_samples_leaf': 1, 'max_depth': 12, 'criterion': 'entropy'}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\andre\\AppData\\Local\\Temp\\ipykernel_5556\\3798158344.py:36: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  random_forestcls_cv.fit(X_train_tfidf, y_train)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data test\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "     negatif       0.55      0.69      0.61        89\n",
      "      netral       0.63      0.64      0.63        91\n",
      "     positif       0.61      0.47      0.53        92\n",
      "\n",
      "    accuracy                           0.60       272\n",
      "   macro avg       0.60      0.60      0.59       272\n",
      "weighted avg       0.60      0.60      0.59       272\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Data Score: 0.8606299212598425\n",
      "Testing Data Score: 0.5955882352941176\n"
     ]
    }
   ],
   "source": [
    "# Parameter eksperimen\n",
    "random_param = {\n",
    "    \"max_depth\": [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12],\n",
    "    'n_estimators': [50, 100, 200],\n",
    "    'min_samples_leaf': [1, 2, 4],\n",
    "    'min_samples_split': [2, 5, 10],\n",
    "    \"criterion\": [\"gini\", \"entropy\"]\n",
    "}\n",
    "\n",
    "# menggunakan randomizedsearch cv untuk cross validation\n",
    "random_forestcls = RandomizedSearchCV(estimator=RandomForestClassifier(random_state=123),\n",
    "                                      param_distributions=random_param,\n",
    "                                      n_iter=50,\n",
    "                                      cv=5,\n",
    "                                      scoring=\"accuracy\",\n",
    "                                      random_state=123)\n",
    "\n",
    "# fit pada data train\n",
    "random_forestcls.fit(X_train_tfidf, np.ravel(y_train))\n",
    "\n",
    "# cek parameter terbaik\n",
    "best_params = random_forestcls.best_params_\n",
    "print(\"Best parameters found: \", best_params)\n",
    "\n",
    "# buat model dengan definisi best params\n",
    "random_forestcls_cv = RandomForestClassifier(\n",
    "    max_depth=best_params[\"max_depth\"],\n",
    "    criterion=best_params[\"criterion\"],\n",
    "    n_estimators=best_params[\"n_estimators\"],\n",
    "    min_samples_leaf=best_params[\"min_samples_leaf\"],\n",
    "    min_samples_split=best_params[\"min_samples_split\"],\n",
    "    random_state=123\n",
    ")\n",
    "\n",
    "# Fit model pada data train\n",
    "random_forestcls_cv.fit(X_train_tfidf, y_train)\n",
    "\n",
    "# cek predict pada data test\n",
    "print(\"Data test\")\n",
    "predict_randomcls_cv = random_forestcls_cv.predict(X_test_tfidf)\n",
    "print(classification_report(y_test, predict_randomcls_cv))\n",
    "\n",
    "# Plot confusion matrix\n",
    "unique_labels = np.unique(y_test)\n",
    "ConfusionMatrixDisplay.from_predictions(y_test, predict_randomcls_cv, display_labels=unique_labels)\n",
    "plt.show()\n",
    "\n",
    "# Evaluasi model\n",
    "print(f\"Training Data Score: {random_forestcls_cv.score(X_train_tfidf, y_train)}\")\n",
    "print(f\"Testing Data Score: {random_forestcls_cv.score(X_test_tfidf, y_test)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "1389010c-2a3b-4a0f-a2d4-f2a5de1e3720",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['random_forest_tuned.joblib']"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Save model\n",
    "dump(random_forestcls_cv, 'random_forest_tuned.joblib')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6058726-2335-4521-8f09-6571bf06e904",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Load model\n",
    "# random_forestcls_cv = load('random_forest_tuned.joblib')"
   ]
  }
 ],
 "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": "3.10.0rc2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}