Prediction of Cancer Types using Patient Gene Expression .ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "99dc41e6",
   "metadata": {},
   "source": [
    "# Project By: Princilla Abena Koranteng"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4e50e71",
   "metadata": {},
   "source": [
    "# PREDICTION OF CANCER TYPE USING PATIENT GENE EXPRESSION"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20d81767",
   "metadata": {},
   "source": [
    "# PROBLEM STATEMENT \n",
    "\n",
    "Cancer has become one of the major factors responsible for global deaths, due to late diagnoses and lack of proper treatment. It involves the abnormal and uncontrolled growth of cells inside the body, which might spread from one place to different parts. Ribonucleic acid (RNA) sequencing can detect the changes occurring inside cells and helps to analyze the transcriptome of gene expression patterns inside RNA. Machine learning techniques can assist in the prediction of cancer at an early stage, if data is available. \n",
    "\n",
    "## Objectives\n",
    "The objective for this project is to build models and classify the different cancer types using RNA-seq gene expression data.\n",
    "\n",
    "For this purpose we implemented supervised models to classify the samples collected using the labels and compare the performance of the respective models."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b05c990e",
   "metadata": {},
   "source": [
    "## Data Description\n",
    "The dataset was collected from UCI Machine Learning Repository. The rows represent the observations gene expressions of patients having different types of tumors (cancer types). The last column contains the cancer types as listed below:\n",
    "- BRCA for Breast Cancer\n",
    "- KIRC for Renal Cancer (Kidney Renal Clear Cell Carcinoma)\n",
    "- LUAD for Lung Cancer (Lung adenocarcinoma)\n",
    "- PRAD for Prostate Cancer (Prostate adenocarcinoma)\n",
    "- COAD fro Colon adenocarcinoma\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "accd2cb3",
   "metadata": {},
   "source": [
    "## Machine Learning Workflow\n",
    "[1. Import Python Libraries](#Import-Python-libraries)\n",
    "\n",
    "[2. Importing data](#Importing-data)\n",
    "\n",
    "[3. Data Exploration & Cleaning](#Data-Exploration-&-Cleaning)\n",
    "\n",
    "[4. Data Preprocesing](#Data-preprocesing)\n",
    "\n",
    "[5. Building & Evaluation of Model](#Building-the-model)\n",
    "\n",
    "   * [-KNeighbours Classifier](#KNeighborsClassifier)\n",
    "   \n",
    "   * [-Logistic Regression](#Logistic-Regression)\n",
    "   \n",
    "   * [-SGDClassifier](#SGDClassifier)\n",
    "   * [-RandomForest Classifier](#RandomForestClassifier)\n",
    "   * [-Decision Tree Classifier](#Decision-Tree-Classifier)\n",
    "   * [-Voting Classifier](#VotingClassifier)  \n",
    "   * [-XGBoost](#XGBoost)\n",
    "   * [-BaggingClassifier](#BaggingClassifier)\n",
    "   * [-AdaBoostClassifier](#AdaBoostClassifier)\n",
    "    \n",
    "[6. Result](#RESULT)   "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91f3dde3",
   "metadata": {},
   "source": [
    "## Import Python libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a0515dec",
   "metadata": {},
   "outputs": [
    {
     "ename": "ModuleNotFoundError",
     "evalue": "No module named 'yellowbrick'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
      "Input \u001b[1;32mIn [15]\u001b[0m, in \u001b[0;36m<cell line: 37>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     33\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01msklearn\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mdiscriminant_analysis\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m LinearDiscriminantAnalysis\n\u001b[0;32m     34\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01msklearn\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mdecomposition\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m PCA\n\u001b[1;32m---> 37\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01myellowbrick\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mclassifier\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m ROCAUC\n",
      "\u001b[1;31mModuleNotFoundError\u001b[0m: No module named 'yellowbrick'"
     ]
    }
   ],
   "source": [
    "#data handling\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "#data visualization\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "#preprocessing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.preprocessing import label_binarize\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "#classification\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "# performance metrics\n",
    "from sklearn.metrics import f1_score,precision_score, recall_score,make_scorer\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import plot_confusion_matrix\n",
    "from sklearn.metrics import roc_curve,auc\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "#Pipeline\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "#Model Selection\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "# Decomposition\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b839234a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting yellowbrick\n",
      "  Downloading yellowbrick-1.5-py3-none-any.whl (282 kB)\n",
      "Requirement already satisfied: matplotlib!=3.0.0,>=2.0.2 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from yellowbrick) (3.5.1)\n",
      "Requirement already satisfied: numpy>=1.16.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from yellowbrick) (1.21.5)\n",
      "Requirement already satisfied: scikit-learn>=1.0.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from yellowbrick) (1.0.2)\n",
      "Requirement already satisfied: cycler>=0.10.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from yellowbrick) (0.11.0)\n",
      "Requirement already satisfied: scipy>=1.0.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from yellowbrick) (1.7.3)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from matplotlib!=3.0.0,>=2.0.2->yellowbrick) (4.25.0)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from matplotlib!=3.0.0,>=2.0.2->yellowbrick) (3.0.4)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from matplotlib!=3.0.0,>=2.0.2->yellowbrick) (1.3.2)\n",
      "Requirement already satisfied: pillow>=6.2.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from matplotlib!=3.0.0,>=2.0.2->yellowbrick) (9.0.1)\n",
      "Requirement already satisfied: packaging>=20.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from matplotlib!=3.0.0,>=2.0.2->yellowbrick) (21.3)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from matplotlib!=3.0.0,>=2.0.2->yellowbrick) (2.8.2)\n",
      "Requirement already satisfied: six>=1.5 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from python-dateutil>=2.7->matplotlib!=3.0.0,>=2.0.2->yellowbrick) (1.16.0)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from scikit-learn>=1.0.0->yellowbrick) (2.2.0)\n",
      "Requirement already satisfied: joblib>=0.11 in c:\\users\\isaac femi\\anaconda3\\lib\\site-packages (from scikit-learn>=1.0.0->yellowbrick) (1.1.0)\n",
      "Installing collected packages: yellowbrick\n",
      "Successfully installed yellowbrick-1.5\n"
     ]
    }
   ],
   "source": [
    "# yellowbrick installation\n",
    "!pip install yellowbrick"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "0f54b621",
   "metadata": {},
   "outputs": [],
   "source": [
    "from yellowbrick.classifier import ROCAUC"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b45976a3",
   "metadata": {},
   "source": [
    "## Importing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "25a5cb29",
   "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>gene_1</th>\n",
       "      <th>gene_2</th>\n",
       "      <th>gene_3</th>\n",
       "      <th>gene_4</th>\n",
       "      <th>gene_5</th>\n",
       "      <th>gene_6</th>\n",
       "      <th>gene_7</th>\n",
       "      <th>gene_8</th>\n",
       "      <th>gene_9</th>\n",
       "      <th>gene_10</th>\n",
       "      <th>...</th>\n",
       "      <th>gene_7992</th>\n",
       "      <th>gene_7993</th>\n",
       "      <th>gene_7994</th>\n",
       "      <th>gene_7995</th>\n",
       "      <th>gene_7996</th>\n",
       "      <th>gene_7997</th>\n",
       "      <th>gene_7998</th>\n",
       "      <th>gene_7999</th>\n",
       "      <th>gene_8000</th>\n",
       "      <th>Cancer_Type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.088413</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.550605</td>\n",
       "      <td>2.815760</td>\n",
       "      <td>...</td>\n",
       "      <td>11.558803</td>\n",
       "      <td>8.881802</td>\n",
       "      <td>6.014840</td>\n",
       "      <td>6.643534</td>\n",
       "      <td>11.740624</td>\n",
       "      <td>7.065012</td>\n",
       "      <td>9.932659</td>\n",
       "      <td>6.928584</td>\n",
       "      <td>2.088413</td>\n",
       "      <td>KIRC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.205955</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.425244</td>\n",
       "      <td>2.354396</td>\n",
       "      <td>...</td>\n",
       "      <td>11.062829</td>\n",
       "      <td>9.032864</td>\n",
       "      <td>5.054193</td>\n",
       "      <td>6.432320</td>\n",
       "      <td>12.104985</td>\n",
       "      <td>7.300746</td>\n",
       "      <td>9.872796</td>\n",
       "      <td>5.039231</td>\n",
       "      <td>2.448002</td>\n",
       "      <td>KIRC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.746646</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.639417</td>\n",
       "      <td>1.657091</td>\n",
       "      <td>...</td>\n",
       "      <td>12.497640</td>\n",
       "      <td>7.198160</td>\n",
       "      <td>0.943434</td>\n",
       "      <td>7.371690</td>\n",
       "      <td>11.202356</td>\n",
       "      <td>8.426588</td>\n",
       "      <td>11.176890</td>\n",
       "      <td>3.119738</td>\n",
       "      <td>4.433988</td>\n",
       "      <td>BRCA</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 8001 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   gene_1  gene_2  gene_3    gene_4  gene_5  gene_6  gene_7  gene_8    gene_9  \\\n",
       "0     0.0     0.0     0.0  2.088413     0.0     0.0     0.0     0.0  0.550605   \n",
       "1     0.0     0.0     0.0  3.205955     0.0     0.0     0.0     0.0  0.425244   \n",
       "2     0.0     0.0     0.0  4.746646     0.0     0.0     0.0     0.0  2.639417   \n",
       "\n",
       "    gene_10  ...  gene_7992  gene_7993  gene_7994  gene_7995  gene_7996  \\\n",
       "0  2.815760  ...  11.558803   8.881802   6.014840   6.643534  11.740624   \n",
       "1  2.354396  ...  11.062829   9.032864   5.054193   6.432320  12.104985   \n",
       "2  1.657091  ...  12.497640   7.198160   0.943434   7.371690  11.202356   \n",
       "\n",
       "   gene_7997  gene_7998  gene_7999  gene_8000  Cancer_Type  \n",
       "0   7.065012   9.932659   6.928584   2.088413         KIRC  \n",
       "1   7.300746   9.872796   5.039231   2.448002         KIRC  \n",
       "2   8.426588  11.176890   3.119738   4.433988         BRCA  \n",
       "\n",
       "[3 rows x 8001 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "#Importing data \n",
    "\n",
    "cancer_data = pd.read_csv('cancer_gene_expression.csv')\n",
    "cancer_data.head(3)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a2f2cfb8",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "e214ad93",
   "metadata": {},
   "source": [
    "\n",
    "## Data Exploration & Cleaning\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "18e48d71",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(801, 8001)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#let's check the number of samples and features\n",
    "#note:the last column contain the labels. it is not considered as a feature\n",
    "\n",
    "cancer_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5ff6d860",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "columns with missing values:0\n"
     ]
    }
   ],
   "source": [
    "#check for missing values\n",
    "datanul=cancer_data.isnull().sum()\n",
    "g=[i for i in datanul if i>0]\n",
    "\n",
    "print('columns with missing values:%d'%len(g))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f04d3638",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Standard Deviation')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(cancer_data.iloc[:,:-1].std())\n",
    "plt.title('Distribution of Standard deviation across features')\n",
    "plt.xlabel('Standard Deviation')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cf1c0103",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Mean')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(cancer_data.iloc[:,:-1].mean())\n",
    "plt.title('Distribution of Mean across features')\n",
    "plt.xlabel('Mean')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b27902e",
   "metadata": {},
   "source": [
    " Based on the graphs above it is noted that although most of the columns have their mean around 0, the standard deviation of the various columns are around 0.5.\n",
    " SInce it is required the standard deviation be 1 for normally distributed data, it is concluded that the data needs to be preprocessed with a scaler before building of models. This will normalize the data and improve effectiveness of the models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "229679e0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d0560c5a",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BRCA    300\n",
      "KIRC    146\n",
      "LUAD    141\n",
      "PRAD    136\n",
      "COAD     78\n",
      "Name: Cancer_Type, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "#checking the counts of the different cancer types that are in the data\n",
    "\n",
    "print(cancer_data['Cancer_Type'].value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "046d64c6",
   "metadata": {},
   "source": [
    " It is noted that there are 5 classes/cancer types. Although the count of BRCA is quite high, it doesn't indicate imbalance that will throw of the effectiveness of the machine learning models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "93990c7a",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot a bar chat to display the class distribution\n",
    "\n",
    "cancer_data['Cancer_Type'].value_counts().plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31adfe33",
   "metadata": {},
   "source": [
    "# Project By: Princilla Abena Koranteng"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a0e165b",
   "metadata": {},
   "source": [
    " \n",
    "## **Data preprocesing** \n",
    "The data is first separated into features and a target variable after which the classes are encoded as numeric type\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "03d4b489",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X\n",
      "   gene_1  gene_2  gene_3    gene_4  gene_5  gene_6  gene_7  gene_8    gene_9  \\\n",
      "0     0.0     0.0     0.0  2.088413     0.0     0.0     0.0     0.0  0.550605   \n",
      "\n",
      "   gene_10  ...  gene_7991  gene_7992  gene_7993  gene_7994  gene_7995  \\\n",
      "0  2.81576  ...   6.883022  11.558803   8.881802    6.01484   6.643534   \n",
      "\n",
      "   gene_7996  gene_7997  gene_7998  gene_7999  gene_8000  \n",
      "0  11.740624   7.065012   9.932659   6.928584   2.088413  \n",
      "\n",
      "[1 rows x 8000 columns]\n",
      "\n",
      "y\n",
      "0    KIRC\n",
      "1    KIRC\n",
      "2    BRCA\n",
      "3    KIRC\n",
      "4    COAD\n",
      "Name: Cancer_Type, dtype: object\n"
     ]
    }
   ],
   "source": [
    "# 1. Separating data into features (X) and the target (y)\n",
    "X = cancer_data.drop(['Cancer_Type'], axis =1)\n",
    "y = cancer_data['Cancer_Type']\n",
    "\n",
    "print('X')\n",
    "print(X.head(1))\n",
    "print()\n",
    "print('y')\n",
    "print(y.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bf36b8de",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "264a01cb",
   "metadata": {},
   "source": [
    "\\\n",
    "**Encode labels**\n",
    "\n",
    "The labels for this data are categorical and we therefore have to convert them to numeric forms. This is referred to as encoding. Machine learning models usually require input data to be in numeric forms, hence we encoding the labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b251f08b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#let's encode target labels (y) with values between 0 and n_classes-1.\n",
    "#encoding will be done using the LabelEncoder\n",
    "label_encoder=LabelEncoder()\n",
    "label_encoder.fit(y)\n",
    "y_encoded=label_encoder.transform(y)\n",
    "labels=label_encoder.classes_\n",
    "classes=np.unique(y_encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4769ffe6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['BRCA', 'COAD', 'KIRC', 'LUAD', 'PRAD'], dtype=object)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8a567de8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 2, 3, 4])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "97640002",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "5f68c4f6",
   "metadata": {},
   "source": [
    "\\\n",
    "**Data Splitting**\\\n",
    "The data will now split into training and test subsets.\n",
    "The training data is initially passed to the machine learning model during fitting. This is to enable the model to identify patterns which can be used to make future predictions.\n",
    "The testing data is used to evaluate the model after the training phase."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6fada994",
   "metadata": {},
   "outputs": [],
   "source": [
    "#split data into training and test sets\n",
    "X_train,X_test,y_train,y_test=train_test_split(X,y_encoded,test_size=0.2,random_state = 42)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a1736a7",
   "metadata": {},
   "source": [
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "57959711",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining function that will run models since a similar process will be done for almost all models\n",
    "def runmodel(model,tuning_params,scorer=make_scorer(f1_score,average='weighted'),n_iter=20):\n",
    "    pca=PCA(random_state=0)\n",
    "    sc=StandardScaler()\n",
    "    pipe=Pipeline(steps=[('sc',sc),('pca',pca),('model',model)])\n",
    "    r_search=RandomizedSearchCV(pipe,tuning_params,n_jobs=-1,verbose=-1,scoring=scorer,cv=4,n_iter=n_iter,random_state=2)\n",
    "    r_search.fit(X_train,y_train)\n",
    "    print(r_search.best_estimator_.get_params())\n",
    "    print(classification_report(y_test,r_search.predict(X_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "502ab5ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining a function for plotting ROC Curves for the various classifiers\n",
    "def plot_roc(model):\n",
    "    features=list(range(27))\n",
    "    plotter=ROCAUC(model,micro=False,classes=['BRCA','COAD','KIRC','LUAD','PRAD'])\n",
    "    plotter.fit(X_train.values,y_train)\n",
    "    plotter.score(X_test.values,y_test)\n",
    "    plotter.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "988beea5",
   "metadata": {},
   "source": [
    "### Building the model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ca03b28",
   "metadata": {},
   "source": [
    "## KNeighborsClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "baae97cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "model=KNeighborsClassifier()\n",
    "tuning_params=dict(model__n_neighbors=list(range(3,13)),pca__n_components=list(range(300)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f3fde932",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'memory': None, 'steps': [('sc', StandardScaler()), ('pca', PCA(n_components=27, random_state=0)), ('model', KNeighborsClassifier(n_neighbors=12))], 'verbose': False, 'sc': StandardScaler(), 'pca': PCA(n_components=27, random_state=0), 'model': KNeighborsClassifier(n_neighbors=12), 'sc__copy': True, 'sc__with_mean': True, 'sc__with_std': True, 'pca__copy': True, 'pca__iterated_power': 'auto', 'pca__n_components': 27, 'pca__random_state': 0, 'pca__svd_solver': 'auto', 'pca__tol': 0.0, 'pca__whiten': False, 'model__algorithm': 'auto', 'model__leaf_size': 30, 'model__metric': 'minkowski', 'model__metric_params': None, 'model__n_jobs': None, 'model__n_neighbors': 12, 'model__p': 2, 'model__weights': 'uniform'}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      1.00      0.98        60\n",
      "           1       1.00      0.94      0.97        16\n",
      "           2       1.00      1.00      1.00        28\n",
      "           3       0.96      0.92      0.94        24\n",
      "           4       1.00      0.97      0.98        33\n",
      "\n",
      "    accuracy                           0.98       161\n",
      "   macro avg       0.98      0.96      0.97       161\n",
      "weighted avg       0.98      0.98      0.98       161\n",
      "\n"
     ]
    }
   ],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "65d7a6fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "knn=KNeighborsClassifier(n_neighbors=12)\n",
    "pca=PCA(n_components=27,random_state=0)\n",
    "sc=StandardScaler()\n",
    "knnpipe=Pipeline(steps=[('sc',sc),('pca',pca),('knn',knn)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "ec920229",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9750363300142004"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knnpipe.fit(X_train.values,y_train)\n",
    "knnf1=f1_score(y_test,knnpipe.predict(X_test.values),average='weighted')\n",
    "knnf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "32bd1347",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roc(knnpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5153115e",
   "metadata": {},
   "source": [
    "### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ed429fc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "model=LogisticRegression(random_state=7,n_jobs=-1,multi_class='multinomial', max_iter=1000,penalty='l2')\n",
    "tuning_params=dict(model__solver=['newton-cg','sag','saga','lbfgs'],\n",
    "                   pca__n_components=list(range(300)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8b82cda5",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d0b7732e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "logreg=LogisticRegression(random_state=7,n_jobs=-1,multi_class='multinomial',penalty='l2',solver='sag', max_iter=1000)\n",
    "pca=PCA(n_components=149,random_state=0)\n",
    "sc=StandardScaler()\n",
    "logregpipe=Pipeline(steps=[('sc',sc),('pca',pca),('logreg',logreg)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6bf39c04",
   "metadata": {},
   "outputs": [],
   "source": [
    "logregpipe.fit(X_train.values,y_train)\n",
    "lrf1=f1_score(y_test,logregpipe.predict(X_test.values),average='weighted')\n",
    "lrf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "95f80737",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(logregpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58c6d53d",
   "metadata": {},
   "source": [
    "### SGDClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "885b7be1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "model=SGDClassifier(n_jobs=-1,random_state=20,average=True)\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),model__loss=['hinge','squared_hinge','perceptron'],\n",
    "                model__alpha=np.logspace(0.001,10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2cf9dc2f",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7bad8d04",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "sgd=SGDClassifier(random_state=7,n_jobs=-1,alpha=68.79407260342448,loss='hinge',average=True)\n",
    "pca=PCA(n_components=154,random_state=0)\n",
    "sc=StandardScaler()\n",
    "sgdpipe=Pipeline(steps=[('sc',sc),('pca',pca),('sgd',sgd)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "47897ace",
   "metadata": {},
   "outputs": [],
   "source": [
    "sgdpipe.fit(X_train.values,y_train)\n",
    "sgdf1=f1_score(y_test,sgdpipe.predict(X_test.values),average='weighted')\n",
    "sgdf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "61bd6c39",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(sgdpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2811365c",
   "metadata": {},
   "source": [
    "### RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2b622ef8",
   "metadata": {},
   "outputs": [],
   "source": [
    "model=RandomForestClassifier(n_jobs=-1,random_state=20,verbose=1)\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),model__n_estimators=[50,70,90,110],\n",
    "                model__criterion=['gini','entropy'],model__max_depth=[2,4,6,8])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aef90bff",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8b2c1ca6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "rforest=RandomForestClassifier(random_state=20,n_jobs=-1,n_estimators=70,criterion='gini',max_depth=6)\n",
    "pca=PCA(n_components=57,random_state=0)\n",
    "sc=StandardScaler()\n",
    "rforestpipe=Pipeline(steps=[('sc',sc),('pca',pca),('rforest',rforest)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6d95f0a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "rforestpipe.fit(X_train.values,y_train)\n",
    "rff1=f1_score(y_test,rforestpipe.predict(X_test.values),average='weighted')\n",
    "rff1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "600824e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(rforestpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3dccc82d",
   "metadata": {},
   "source": [
    "### Decision Tree Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "79519295",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "model=DecisionTreeClassifier(random_state=0)\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),\n",
    "                model__criterion=['gini','entropy'],model__max_depth=list(range(2,8)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "79812049",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8988d6f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "dtree=DecisionTreeClassifier(random_state=0,criterion='entropy',max_depth=6)\n",
    "pca=PCA(n_components=179,random_state=0)\n",
    "sc=StandardScaler()\n",
    "dtreepipe=Pipeline(steps=[('sc',sc),('pca',pca),('dtree',dtree)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2da90e4c",
   "metadata": {},
   "outputs": [],
   "source": [
    "dtreepipe.fit(X_train.values,y_train)\n",
    "dtf1=f1_score(y_test,dtreepipe.predict(X_test.values),average='weighted')\n",
    "dtf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9e19bdad",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(dtreepipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08fbc80a",
   "metadata": {},
   "source": [
    "### VotingClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "505c1af1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier as KNN\n",
    "from sklearn.ensemble import VotingClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56b8ddbd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate individual classifiers\n",
    "lr = LogisticRegression(random_state=42, max_iter=1000)\n",
    "knn = KNN()\n",
    "dtr = DecisionTreeClassifier(random_state=42)\n",
    "# Define a list called classifier that contains the tuples (classifier_name, classifier)\n",
    "classifiers = [('Logistic Regression', lr),\n",
    "('K Nearest Neighbours', knn),\n",
    "('Classification Tree', dtr)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "77c405c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Iterate over the defined list of tuples containing the classifiers\n",
    "for clf_name, clf in classifiers:\n",
    "        #fit clf to the training set\n",
    "        clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f002f233",
   "metadata": {},
   "outputs": [],
   "source": [
    "model=VotingClassifier(estimators=classifiers,n_jobs=-1)\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),\n",
    "                model__voting=['hard','soft'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1539e097",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8e12873a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "vc=VotingClassifier(estimators=classifiers,n_jobs=-1,voting='soft')\n",
    "pca=PCA(n_components=245,random_state=0)\n",
    "sc=StandardScaler()\n",
    "vcpipe=Pipeline(steps=[('sc',sc),('pca',pca),('vc',vc)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e335a3ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "vcpipe.fit(X_train.values,y_train)\n",
    "vcf1=f1_score(y_test,vcpipe.predict(X_test.values),average='weighted')\n",
    "vcf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca298745",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(vcpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3fa010b",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce6e2ed6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Installing XGBoost\n",
    "!pip install xgboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a475f2ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import XGBoost\n",
    "import xgboost as xgb\n",
    "model=xgb.XGBClassifier(random_state=5,num_class=5,as_pandas=True,early_stopping_round=10,num_boost_round=10,\n",
    "                        objective='multi:softmax')\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),model__colsample_bytree=[0.2,0.4,0.6,0.8],\n",
    "                   model__n_estimators=[30,50,70,90],model__max_depth=[2,4,6,8,10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7e984007",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "732490ae",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "xgb=xgb.XGBClassifier(random_state=5,num_class=5,objective='multi:softmax',colsample_bytree=0.4,n_estimators=70,max_depth=2)\n",
    "pca=PCA(n_components=106,random_state=0)\n",
    "sc=StandardScaler()\n",
    "xgbpipe=Pipeline(steps=[('sc',sc),('pca',pca),('xgb',xgb)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "000ebf40",
   "metadata": {},
   "outputs": [],
   "source": [
    "xgbpipe.fit(X_train.values,y_train)\n",
    "xgbf1=f1_score(y_test,xgbpipe.predict(X_test.values),average='weighted')\n",
    "xgbf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e114cb02",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(xgbpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8fafef56",
   "metadata": {},
   "source": [
    "### BaggingClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a4828735",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import BaggingClassifier\n",
    "\n",
    "# instantiating a DecisionTreeClassifier as a base model\n",
    "dt = DecisionTreeClassifier(max_depth = 4, random_state =123)\n",
    "model = BaggingClassifier(base_estimator=dt,n_jobs = -1)\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),model__max_samples=[0.2,0.4,0.6,0.8],\n",
    "                   model__n_estimators=list(range(400)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f74a6b79",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "255fba84",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "bag=BaggingClassifier(base_estimator=dt,n_jobs=-1,n_estimators=114,max_samples=0.2)\n",
    "pca=PCA(n_components=12,random_state=0)\n",
    "sc=StandardScaler()\n",
    "bagpipe=Pipeline(steps=[('sc',sc),('pca',pca),('bag',bag)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2ee604f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "bagpipe.fit(X_train.values,y_train)\n",
    "bagf1=f1_score(y_test,bagpipe.predict(X_test.values),average='weighted')\n",
    "bagf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "45b7fa6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(bagpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d92d9902",
   "metadata": {},
   "source": [
    "### SVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84136f5c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC \n",
    "model = SVC(random_state=9,decision_function_shape='ovo')\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),model__C=[0.2,0.4,0.6,0.8],\n",
    "                   model__kernel=['rbf','poly','sigmoid'],model__degree=list(range(3,5)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dd182b0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a9c8c7c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "svc=SVC(random_state=9,decision_function_shape='ovo',degree=3,kernel='sigmoid',C=0.4)\n",
    "pca=PCA(n_components=262,random_state=0)\n",
    "sc=StandardScaler()\n",
    "svcpipe=Pipeline(steps=[('sc',sc),('pca',pca),('svc',svc)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "03aaece0",
   "metadata": {},
   "outputs": [],
   "source": [
    "svcpipe.fit(X_train.values,y_train)\n",
    "svcf1=f1_score(y_test,svcpipe.predict(X_test.values),average='weighted')\n",
    "svcf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cf3b1560",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(svcpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65c52291",
   "metadata": {},
   "source": [
    "### AdaBoostClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "674c3a1a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "model=AdaBoostClassifier(random_state=132)\n",
    "tuning_params=dict(pca__n_components=list(range(3,300,1)),model__learning_rate=[0.2,0.4,0.6,0.8,1.0],\n",
    "                   model__n_estimators=list(range(350)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e75e3d82",
   "metadata": {},
   "outputs": [],
   "source": [
    "runmodel(model,tuning_params,n_iter=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1fd78dae",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on result of the Randomize search, the optimized hyperparameters are used to create the model\n",
    "ada=AdaBoostClassifier(random_state=11,learning_rate=0.4,n_estimators=334)\n",
    "pca=PCA(n_components=334,random_state=0)\n",
    "sc=StandardScaler()\n",
    "adapipe=Pipeline(steps=[('sc',sc),('pca',pca),('ada',ada)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a3bc0021",
   "metadata": {},
   "outputs": [],
   "source": [
    "adapipe.fit(X_train.values,y_train)\n",
    "adaf1=f1_score(y_test,adapipe.predict(X_test.values),average='weighted')\n",
    "adaf1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1cb9c489",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_roc(adapipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1dc4a1a",
   "metadata": {},
   "source": [
    "# RESULT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b6fe4ebc",
   "metadata": {},
   "outputs": [],
   "source": [
    "f1_dict={\"Model\":['AdaBoost','KNeighbors Classifier','LogisticRegression','SGDClassifier','RandomForestClassifier',\n",
    "                 'DecisionTreeClassifier','VotingClassifier','XGBoostClassifier','BaggingClassifier','SVC'],\n",
    "        \"F1_Scores\":[adaf1,knnf1,lrf1,sgdf1,rff1,dtf1,vcf1,xgbf1,bagf1,svcf1]}\n",
    "f1_df=pd.DataFrame.from_dict(f1_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ae3daa1a",
   "metadata": {},
   "outputs": [],
   "source": [
    "f1_df.plot.bar(x='Model',y='F1_Scores')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1980494",
   "metadata": {},
   "source": [
    "Based on the f1 scores of the various models, the top two models are LogisticRegression(logregpipe) with f1_score of 1.0 and XGBoost(xgbpipe) with f1 score of 0.993817\n",
    "\n",
    "XGBoost is however closely followed by SVC(svcpipe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3be2b966",
   "metadata": {},
   "source": [
    "# Project By: Princilla Abena Koranteng"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed9e2868",
   "metadata": {},
   "source": [
    "# THANK YOU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba4e3799",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4ead6428",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}