ML_methods_Final.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b588753b",
   "metadata": {
    "id": "b588753b"
   },
   "outputs": [],
   "source": [
    "import time\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import classification_report,confusion_matrix, f1_score, accuracy_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from matplotlib.pyplot import figure\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "28178a07",
   "metadata": {
    "id": "28178a07"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(506, 9318)\n",
      "(506, 1)\n"
     ]
    }
   ],
   "source": [
    "#Load in the data from Dexter, so we both use the same train/test datasets\n",
    "from sklearn.model_selection import train_test_split\n",
    "dataset_train = pd.read_csv('./oversampled_train.csv', index_col=0)\n",
    "dataset_test = pd.read_csv('./x_test.csv', index_col=0)\n",
    "oversampled_data = pd.read_csv('./oversampled_train.csv', index_col=0)\n",
    "\n",
    "dataset_train = dataset_train.replace({'Index' : { False : 'MSI-L', True : 'MSI-H'}})\n",
    "dataset_test = dataset_test.replace({'Index' : { False : 'MSI-L', True : 'MSI-H'}})\n",
    "oversampled_data = oversampled_data.replace({'Index' : { False : 'MSI-L', True : 'MSI-H'}})\n",
    "\n",
    "x_train = dataset_train.iloc[:, dataset_train.columns != 'Index']\n",
    "y_train = dataset_train.iloc[:, [0]]\n",
    "\n",
    "x_test = dataset_test.iloc[:, dataset_test.columns != 'Index']\n",
    "y_test = dataset_test.iloc[:, [0]]\n",
    "\n",
    "x_train_oversampled = oversampled_data.iloc[:, oversampled_data.columns != 'Index']\n",
    "y_train_oversampled = oversampled_data.iloc[:, [0]]\n",
    "\n",
    "print(x_train_oversampled.shape)\n",
    "print(y_train_oversampled.shape)\n",
    "\n",
    "# x_train_oversampled, x_val_oversampled, y_train_oversampled, y_val_oversampled = train_test_split(x_train_oversampled, y_train_oversampled, test_size=0.1, stratify=y_train_oversampled)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "2b5c2aa7",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(506, 9318)\n",
      "(506, 1)\n",
      "(506, 9318)\n",
      "(506, 1)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape)\n",
    "print(y_train.shape)\n",
    "print(x_train_oversampled.shape)\n",
    "print(y_train_oversampled.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7a70a9d3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# rf = RandomForestClassifier()\n",
    "# param_grid = {\n",
    "#     'n_estimators': [50, 100, 150, 500],\n",
    "#     'max_depth': [2, 5, 7, 9]\n",
    "# }\n",
    "# from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# grid_rf = GridSearchCV(rf, param_grid, cv=5)\n",
    "# grid_rf.fit(x_train, y_train.values.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "acb2aec5",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# print(grid_rf.best_estimator_)\n",
    "# print(grid_rf.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "bb4867f7",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(131, 9318)\n",
      "(506, 9318)\n",
      "(131, 500)\n",
      "(506, 500)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainedforest = RandomForestClassifier(n_estimators=900).fit(x_train,y_train.values.ravel())\n",
    "#get the top 500 features\n",
    "feat_importances = pd.Series(trainedforest.feature_importances_, index= x_train.columns)\n",
    "feat_importances.nlargest(500).plot(kind='barh')\n",
    "top_features = feat_importances.nlargest(500).index\n",
    "#print(top_features)\n",
    "# print(top_features)\n",
    "top_features = list(top_features)\n",
    "#transpose x_train + x_test to get the features as row names again.\n",
    "x_train = x_train.T\n",
    "x_test = x_test.T\n",
    "\n",
    "top_features_list = []\n",
    "x_train_features_list = []\n",
    "x_test_list = []\n",
    "\n",
    "x_train_index = x_train.index\n",
    "x_test_index = x_test.index\n",
    "\n",
    "for x in top_features:\n",
    "    top_features_list.append(x)\n",
    "for x in x_train_index:\n",
    "     x_train_features_list.append(x)\n",
    "for x in x_test_index:\n",
    "     x_test_list.append(x)\n",
    "\n",
    "difference_features = set(x_train_features_list).difference(set(top_features_list))\n",
    "difference_features_list = list(difference_features)\n",
    "\n",
    "difference_features_test = set(x_test_list).difference(set(top_features_list))\n",
    "difference_features_list_test = list(difference_features_test)\n",
    "\n",
    "#remove features from the trainingsset\n",
    "x_train_tf = x_train[~x_train.index.isin(difference_features_list)]\n",
    "x_test_tf = x_test[~x_test.index.isin(difference_features_list_test)]\n",
    "\n",
    "\n",
    "#transpose x_train_tf + x_test_tf\n",
    "x_train_tf = x_train_tf.T\n",
    "x_test_tf = x_test_tf.T\n",
    "\n",
    "#transpose x_train + x_test back to original\n",
    "x_test = x_test.T\n",
    "# x_val = x_val.T\n",
    "x_train = x_train.T\n",
    "# x_train2 = x_train2.T\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "x_train_tf = StandardScaler().fit_transform(x_train_tf)\n",
    "x_test_tf = StandardScaler().fit_transform(x_test_tf)\n",
    "print(x_test.shape)\n",
    "print(x_train.shape)\n",
    "\n",
    "print(x_test_tf.shape)\n",
    "print(x_train_tf.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "6a12d66b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Write top 500 genes to a file\n",
    "with open('top_features.txt', 'w') as f:\n",
    "    for x in top_features:\n",
    "        f.write(x+\"\\n\")  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "06d91152",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRR15      0.012938\n",
      "CRCP       0.011682\n",
      "PLAGL2     0.011587\n",
      "HSPH1      0.010211\n",
      "AXIN2      0.008818\n",
      "             ...   \n",
      "SHROOM1    0.000282\n",
      "SLC25A4    0.000281\n",
      "RPL30      0.000281\n",
      "SLC35F5    0.000280\n",
      "NCAPH2     0.000279\n",
      "Length: 500, dtype: float64\n",
      "NCAPH2\n"
     ]
    }
   ],
   "source": [
    "#Write top 500 genes to a file with importance\n",
    "from pathlib import Path  \n",
    "filepath = Path('top_features_and_importance.csv')  \n",
    "filepath.parent.mkdir(parents=True, exist_ok=True)  \n",
    "Top_500 = feat_importances.nlargest(500)\n",
    "Top_500.to_string()\n",
    "print(Top_500)\n",
    "\n",
    "Top_500 = list(zip(Top_500,top_features))\n",
    "Top_500 = pd.DataFrame(Top_500,columns=[\"Importance\",\"Feature_Name\"])\n",
    "print(x)\n",
    "Top_500.to_csv('Top_features.csv', index=False, header=False)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bc3aa211",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9901184236070666\n",
      "[[253   0]\n",
      " [  7 246]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9541984732824428\n"
     ]
    }
   ],
   "source": [
    "#random forest with oversampling only\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.datasets import make_classification\n",
    "from sklearn.metrics import RocCurveDisplay\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.model_selection import cross_val_predict\n",
    "#x_train_oversampled and y_train_oversampled are made with the oversampled dataset from Dexter! I just loaded them in!!!!\n",
    "\n",
    "rf = RandomForestClassifier(n_estimators=900, max_depth=6)\n",
    "rf.fit(x_train_oversampled, y_train_oversampled.values.ravel())\n",
    "print(np.mean(cross_val_score(rf, x_train_oversampled, y_train_oversampled.values.ravel(), cv=5)))\n",
    "y_pred_rf = cross_val_predict(rf, x_train_oversampled, y_train_oversampled.values.ravel(), cv=5)\n",
    "conf_mat_rf = confusion_matrix(y_train_oversampled.values.ravel(), y_pred_rf)\n",
    "print(conf_mat_rf)\n",
    "rf_disp = RocCurveDisplay.from_estimator(rf, x_test, y_test,pos_label='MSI-H')\n",
    "plt.show()\n",
    "prediction = rf.predict(x_test)\n",
    "print(accuracy_score(y_test, prediction))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "68569e5c",
   "metadata": {
    "id": "68569e5c",
    "outputId": "75147ef5-fdb2-414b-ab5a-f44b218f98ff",
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# ax = plt.gca()\n",
    "# rf_disp = RocCurveDisplay.from_estimator(rf, x_val, y_val, ax=ax, alpha=0.8)\n",
    "# rf_disp.plot(ax=ax, alpha=0.8)\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "45bd57f2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9822170452339352\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9541984732824428\n"
     ]
    }
   ],
   "source": [
    "#Logistic regression\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "lg = LogisticRegression(random_state=0, max_iter=1000, solver='lbfgs').fit(x_train_oversampled, y_train_oversampled.values.ravel())\n",
    "print(np.mean(cross_val_score(lg, x_train_oversampled, y_train_oversampled.values.ravel(), cv=5)))\n",
    "lg_disp = RocCurveDisplay.from_estimator(lg, x_test, y_test,pos_label='MSI-H')\n",
    "plt.show()\n",
    "\n",
    "prediction = rf.predict(x_test)\n",
    "print(accuracy_score(y_test, prediction))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "351fb189",
   "metadata": {
    "id": "351fb189",
    "outputId": "9745e497-d44c-4227-f920-19fa3360f7b4",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.998019801980198\n",
      "[[253   0]\n",
      " [  1 252]]\n",
      "0.9007633587786259\n"
     ]
    }
   ],
   "source": [
    "#SVM\n",
    "import numpy as np\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.svm import SVC\n",
    "from sklearn import svm\n",
    "from sklearn.model_selection import cross_val_predict\n",
    "svclassifier = SVC(kernel='rbf')\n",
    "svclassifier.fit(x_train_tf, y_train.values.ravel())\n",
    "print(np.mean(cross_val_score(svclassifier, x_train_tf, y_train.values.ravel(), cv=5)))\n",
    "y_pred = cross_val_predict(svclassifier, x_train_tf, y_train.values.ravel(), cv=5)\n",
    "conf_mat = confusion_matrix(y_train.values.ravel(), y_pred)\n",
    "print(conf_mat)\n",
    "\n",
    "prediction = svclassifier.predict(x_test_tf)\n",
    "print(accuracy_score(y_test, prediction))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "d27e32f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(131, 500)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import RocCurveDisplay\n",
    "print(x_test_tf.shape)\n",
    "svc_disp = RocCurveDisplay.from_estimator(svclassifier, x_test_tf, y_test,pos_label='MSI-H')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f2867d49",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}