ALL_Detailed_final_group_10.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "7f4e343f",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pylab as pl\n",
    "import numpy as np\n",
    "from sklearn.neighbors import KNeighborsClassifier #importing classifier analysing the data by K-nearest neighboring method"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8dc0a91",
   "metadata": {},
   "source": [
    "# Importing X(data.csv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c90e4142",
   "metadata": {},
   "outputs": [],
   "source": [
    "X=np.array(np.loadtxt(\"data.csv\",delimiter=\",\",dtype=np.str)) #importing data.csv file into an array\n",
    "\n",
    "samples_X=X[1:,0] #Extract sample name of the dataset\n",
    "X=np.float_(X[1:,1:])#taking only the features without taking the sample & features names\n",
    "\n",
    "samples_X_strip=samples_X #sample_X_strip is to define the elements of samples_X remain after removing the missing values\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "844632de",
   "metadata": {},
   "source": [
    "# Checking X (sample number, feature number and missing values) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "14ba435e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples of X: 12060\t\t\n",
      "Number of features of X:801\t\t\n",
      "Missing Values: 0\t\t\n",
      "Percentage of Missing values:0\t\t\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of samples of X: %d\\t\\t\"%X.shape[0])\n",
    "print(\"Number of features of X:%d\\t\\t\"%X.shape[1])\n",
    "\n",
    "\n",
    "missing_values=(np.isnan(X)).sum() #isnan() function, to check the missing values\n",
    "print(\"Missing Values: %d\\t\\t\"%missing_values)\n",
    "percentage_missing=(missing_values/X.size)*100\n",
    "print(\"Percentage of Missing values:%d\\t\\t\"%percentage_missing)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72e56355",
   "metadata": {},
   "source": [
    "# Importing y(toxicity_labels.csv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "d1d56f1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "y=[]  #starting point of y variable \n",
    "samples_y=[]  #starting point of the variable that defines the sample names\n",
    "\n",
    "\n",
    "f=open(\"toxicity_labels.csv\",\"r\") #opens file toxicity_labels.csv on read only mode\n",
    "for line in f:\n",
    "    sv=line.strip().split(\",\")\n",
    "    samples_y.append(str(sv[0]).replace(\"NA\",\"NaN\")) #replacing NA values into Not a Number(NaN) values in regard to represent the missing values in the data\n",
    "    \n",
    "    \n",
    "    row=[]\n",
    "    for element in sv:\n",
    "        if element==0:\n",
    "            continue\n",
    "        row.append(str(element).replace(\"NA\",\"NaN\"))\n",
    "    y.append(row)\n",
    "    \n",
    "f.close() #closes toxicity_labels.csv\n",
    "y=np.array(y) #Converting y tuple into an array\n",
    "\n",
    "y=np.float_(y[1:,1]) #extracting all samples with only our target NR.AhR label\n",
    "\n",
    "samples_y=np.array(samples_y[1:])  ##assigning samples_y values into samples_y array\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e75b4659",
   "metadata": {},
   "source": [
    "#### Removing missing values from X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "885fb1cf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8441, 801)\n"
     ]
    }
   ],
   "source": [
    "X_strip=np.delete(X,sample_indices_to_remove,axis=0)\n",
    "samples_X_strip=np.delete(samples_X_strip,sample_indices_to_remove,axis=0)\n",
    "print(X_strip.shape) #number samples and features of y after removing the missing values defined by X_strip"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c5ddc3f8",
   "metadata": {},
   "source": [
    "# Remaining data Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "8d9b7cf1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler #here,StandardScaler is for data normalization which works by removing the mean(mean=0) and scaling each feature into unit variance (Standard deviation=1)\n",
    "scaler=StandardScaler()\n",
    "X_strip=scaler.fit_transform(X_strip) #fit is to calculating the mean and varience of each features and transform function returns the dataset produced by it\n",
    "                                      #fit_transform is used together to improve the efficiency of the model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5af6c34a",
   "metadata": {},
   "source": [
    "Generating figure to visualizing data normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e203483c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pl.figure(figsize=(10,6))\n",
    "ax=pl.subplot(111)\n",
    "ax.boxplot(X_strip)\n",
    "ax.grid(color=\"#CCCCCC\")\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "pl.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4f4684a",
   "metadata": {},
   "source": [
    "# Data Training and Testing\n",
    "\n",
    "- spliting the X_strip and y_strip into train and test data by defining X_strip_test, X_strip_train, y_strip_test, y_strip_train\n",
    "\n",
    "- here, we have maintained the test size =0.2, that means we have 20% validation data and 80% training data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "f80d275c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split #train_test_split is for splitting the data arrays into train dataset and test dataset \n",
    "X_strip_train,X_strip_test,y_train,y_test=train_test_split(X_strip,y,test_size=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73319c1b",
   "metadata": {},
   "source": [
    "# Dimention Reduction and Principal Component Analysis of the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "466bebf2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First 0.00 account for 0.00% of the total variance\n",
      "First 1.00 account for 31.19% of the total variance\n",
      "First 2.00 account for 37.92% of the total variance\n",
      "First 3.00 account for 44.17% of the total variance\n",
      "First 4.00 account for 49.32% of the total variance\n",
      "First 5.00 account for 53.69% of the total variance\n",
      "First 6.00 account for 57.06% of the total variance\n",
      "First 7.00 account for 59.62% of the total variance\n",
      "First 8.00 account for 61.94% of the total variance\n",
      "First 9.00 account for 64.02% of the total variance\n",
      "First 10.00 account for 65.63% of the total variance\n",
      "First 11.00 account for 67.16% of the total variance\n",
      "First 12.00 account for 68.53% of the total variance\n",
      "First 13.00 account for 69.82% of the total variance\n",
      "First 14.00 account for 71.00% of the total variance\n",
      "First 15.00 account for 72.09% of the total variance\n",
      "First 16.00 account for 73.07% of the total variance\n",
      "First 17.00 account for 73.93% of the total variance\n",
      "First 18.00 account for 74.72% of the total variance\n",
      "First 19.00 account for 75.44% of the total variance\n",
      "First 20.00 account for 76.14% of the total variance\n",
      "First 21.00 account for 76.79% of the total variance\n",
      "First 22.00 account for 77.42% of the total variance\n",
      "First 23.00 account for 78.04% of the total variance\n",
      "First 24.00 account for 78.64% of the total variance\n",
      "First 25.00 account for 79.22% of the total variance\n",
      "First 26.00 account for 79.79% of the total variance\n",
      "First 27.00 account for 80.33% of the total variance\n",
      "First 28.00 account for 80.86% of the total variance\n",
      "First 29.00 account for 81.37% of the total variance\n",
      "First 30.00 account for 81.87% of the total variance\n",
      "First 31.00 account for 82.35% of the total variance\n",
      "First 32.00 account for 82.79% of the total variance\n",
      "First 33.00 account for 83.21% of the total variance\n",
      "First 34.00 account for 83.61% of the total variance\n",
      "First 35.00 account for 84.00% of the total variance\n",
      "First 36.00 account for 84.38% of the total variance\n",
      "First 37.00 account for 84.74% of the total variance\n",
      "First 38.00 account for 85.08% of the total variance\n",
      "First 39.00 account for 85.41% of the total variance\n",
      "First 40.00 account for 85.74% of the total variance\n",
      "First 41.00 account for 86.05% of the total variance\n",
      "First 42.00 account for 86.36% of the total variance\n",
      "First 43.00 account for 86.65% of the total variance\n",
      "First 44.00 account for 86.94% of the total variance\n",
      "First 45.00 account for 87.22% of the total variance\n",
      "First 46.00 account for 87.47% of the total variance\n",
      "First 47.00 account for 87.73% of the total variance\n",
      "First 48.00 account for 87.97% of the total variance\n",
      "First 49.00 account for 88.21% of the total variance\n",
      "First 50.00 account for 88.45% of the total variance\n",
      "First 51.00 account for 88.67% of the total variance\n",
      "First 52.00 account for 88.89% of the total variance\n",
      "First 53.00 account for 89.10% of the total variance\n",
      "First 54.00 account for 89.30% of the total variance\n",
      "First 55.00 account for 89.50% of the total variance\n",
      "First 56.00 account for 89.68% of the total variance\n",
      "First 57.00 account for 89.87% of the total variance\n",
      "First 58.00 account for 90.04% of the total variance\n"
     ]
    }
   ],
   "source": [
    "#computation of the covariance metrix(C) into standarized data\n",
    "n=X_strip.shape[0] #samples of X_strip\n",
    "C=1/(n-1)*np.dot(X_strip_train.T,X_strip_train)\n",
    "\n",
    "#eigendecompose the matrix C into its eigenvalues and eigenvectors\n",
    "import numpy.linalg as linalg #linalg function is used to compute eigenvalues and eigenvectors of the squared array\n",
    "d,V=linalg.eig(C) #eigenvalues d and eigenvectors V\n",
    "\n",
    "\n",
    "ind=np.argsort(d)[::-1] #sorting the eigenvalues in the decreasing order\n",
    "# resorting the columns of the eigenvector matrix V using the indices from the sorting of the eigenvalues\n",
    "d=d[ind]\n",
    "V=V[:,ind]\n",
    "\n",
    "\n",
    "i=0\n",
    "vaMax=0.9 #maximum variance 0.9 (90%)\n",
    "ratios_variance_explained=d/d.sum() #computing ratio variance by eigenvalues\n",
    "va=ratios_variance_explained[0:i].sum()\n",
    "\n",
    "while va<=vaMax:\n",
    "    va=ratios_variance_explained[0:i].sum()\n",
    "    print(\"First %.2f account for %.2f%% of the total variance\"%(i,va*100))\n",
    "    i+=1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "345861b5",
   "metadata": {},
   "source": [
    "With the eigenvector of \\begin{align} \\textbf{V}_r  = 60\\end{align},\n",
    "\n",
    "The reduced X is computed by:\n",
    "\n",
    "\\begin{align}\n",
    "\\textbf{X}_r = \\textbf{X_strip} \\textbf{V}_r\n",
    "\\end{align}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c71556ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "Xr=np.dot(X_strip,V[:,0:60]) #computing lower dimentional representation of out input data matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60497d6c",
   "metadata": {},
   "source": [
    "creating a bar chart to visualize the ratio variance explained for each Principal Component"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ae8959c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pl.plot(0,1)\n",
    "xticks=np.arange(1,len(ratios_variance_explained)+1)\n",
    "pl.bar(xticks,ratios_variance_explained,width=20)\n",
    "pl.plot(xticks,np.cumsum(ratios_variance_explained),color=\"tab:red\")\n",
    "#pl.subplot(111)\n",
    "#ax.grid(color=\"#CCCCCC\")\n",
    "pl.ylabel(\"Variance Explained\")\n",
    "pl.xlabel(\"PC\")\n",
    "pl.xticks(xticks)\n",
    "pl.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1192986",
   "metadata": {},
   "source": [
    "# Model Evalution"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3466334b",
   "metadata": {},
   "source": [
    "# Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "3b934a1c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "/home/pinky/miniconda3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression\n",
      "ROC-AUC:\t0.84 (+-0.02)\n",
      "Accuracy:\t0.94 (+-0.00)\n",
      "Recall:\t\t0.70 (+-0.04)\n",
      "Precision:\t0.79 (+-0.03)\n",
      "MCC:\t\t0.71 (+-0.02)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 700x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn import metrics  #metrics is used to monitor and measure the performance of the model\n",
    "from sklearn.model_selection import StratifiedKFold  #StratifiedKFold is for cross validation by spliting data into train/test\n",
    "from sklearn.linear_model import LogisticRegression #importing logisticregression for evaluting the model through logisticregression approach\n",
    "\n",
    "#10 fold cross validation by stratifiedKFold\n",
    "cv = StratifiedKFold(n_splits=10,shuffle=True,random_state=42)\n",
    "#n_splits=10: means 10 fold cross validation, shuffle= True: means it will use each n_split parts as test set,\n",
    "#random_state: for generating reproducible splits\n",
    "\n",
    "list_fpr_lr = [] #list for false positive result of logistic regression\n",
    "list_tpr_lr = [] #list for true positive result of logistic regression\n",
    "auc_values_lr = []  #list for area under the curve of logistic regression \n",
    "precision_values = [] #list for precision values of logistic regression\n",
    "recall_values = [] #list for recall values of logistic regression\n",
    "accuracy_values = [] #list for accuracy values of logistic regression\n",
    "mcc_values = [] #list for Matthews correlation coefficient values of logistic regression\n",
    "\n",
    "\n",
    "#training data\n",
    "#spliting X_strip and y_strip data for the logistic regression baesd on the cross validation by the startfied Kfold\n",
    "for train_index, val_index in cv.split(X_strip,y_strip):\n",
    "    #Split data\n",
    "    X_strip_train = X_strip[train_index]\n",
    "    X_strip_val = X_strip[val_index]\n",
    "    y_train = y[train_index]\n",
    "    y_val = y[val_index]\n",
    "    \n",
    "    #Train Model with simple logistic regression without any penalty\n",
    "    model = LogisticRegression(penalty=\"none\",max_iter=5000,solver=\"lbfgs\")\n",
    "#here penalty is none, maximum iteration for regulating the data optimization, solver=\"lbfgs\" store last few updates to save the memory\n",
    "    model.fit(X_strip_train,y_train) #train the model\n",
    "    \n",
    "    #compute metrics for ROC Curve\n",
    "    scores = model.decision_function(X_strip_val)\n",
    "    fpr, tpr, thresholds = metrics.roc_curve(y_val, scores, drop_intermediate=False)\n",
    "    \n",
    "    list_fpr_lr.append(fpr)\n",
    "    list_tpr_lr.append(tpr)\n",
    "    \n",
    "    y_pred = model.predict(X_strip_val)\n",
    "    auc_values_lr.append(metrics.roc_auc_score(y_val, y_pred))\n",
    "    accuracy_values.append(metrics.accuracy_score(y_val, y_pred))\n",
    "    recall_values.append(metrics.recall_score(y_val, y_pred))\n",
    "    precision_values.append(metrics.precision_score(y_val, y_pred))\n",
    "    mcc_values.append(metrics.matthews_corrcoef(y_val, y_pred))\n",
    "\n",
    "\n",
    "#print results\n",
    "print(\"Logistic Regression\")\n",
    "print(\"ROC-AUC:\\t%.2f (+-%.2f)\" % (np.mean(auc_values_lr),np.std(auc_values_lr)))\n",
    "print(\"Accuracy:\\t%.2f (+-%.2f)\" % (np.mean(accuracy_values),np.std(accuracy_values)))\n",
    "print(\"Recall:\\t\\t%.2f (+-%.2f)\" % (np.mean(recall_values),np.std(recall_values)))\n",
    "print(\"Precision:\\t%.2f (+-%.2f)\" % (np.mean(precision_values),np.std(precision_values)))\n",
    "print(\"MCC:\\t\\t%.2f (+-%.2f)\" % (np.mean(mcc_values),np.std(mcc_values)))\n",
    "\n",
    "#logistic Regression with regularization\n",
    "#generate figure\n",
    "fig = pl.figure(figsize=(7,7)) #axis range\n",
    "ax = fig.add_subplot(111) #plotting subplot\n",
    "\n",
    "#plot roc curve for all 10 foldes\n",
    "pl.plot(fpr, tpr,label=\"Logistic Regression 10 folds (AUC) = %.2f (+- %.2f)\" % (np.mean(auc_values_lr),np.std(auc_values_lr)))\n",
    "\n",
    "\n",
    "\n",
    "ax.plot([0,1], [0,1], color=\"grey\",label=\"Random Classifier\",linestyle=\"--\")\n",
    "#Set axis labels\n",
    "ax.set_xlabel(\"False Positive Rate\")\n",
    "ax.set_ylabel(\"True Positive Rate\")\n",
    "#Set axis limits\n",
    "ax.set_ylim(0,1)\n",
    "ax.set_xlim(0,1)\n",
    "#show grid in grey and set top and right axis to invisible\n",
    "ax.grid(color=\"#CCCCCC\")\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "pl.legend()\n",
    "pl.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "230e4946",
   "metadata": {},
   "source": [
    "# K-nearest neighboring"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "a7b35add",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNearst Neighbors\n",
      "Average k: 2.80 (+- 1.47)\n",
      "Average Accuracy (val): 0.90 (+- 0.00)\n",
      "Average Accuracy (test): 0.91 (+- 0.00)\n",
      "Average Precision (test): 0.71 (+- 0.06)\n",
      "Average Recall (test): 0.39 (+- 0.11)\n",
      "Average ROC-AUC (test): 0.68 (+- 0.05)\n",
      "Average MCC (test): 0.48 (+- 0.04)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 700x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import StratifiedKFold  #StratifiedKFold is for cross validation by spliting data into train/test\n",
    "from sklearn.model_selection import GridSearchCV #GridSearchCV use the cross validation method to evaluate the model by trying all the combination of the values passed in the dictonary\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import metrics #metrics is used to monitor and measure the performance of the model\n",
    " \n",
    "\n",
    "#spliting train/testing data into 5 folds\n",
    "kf = StratifiedKFold(n_splits=5,random_state=42,shuffle=True)\n",
    "#n_splits=5: means 5 fold cross validation, shuffle= True: means it will use each n_split parts as test set,\n",
    "#random_state: for generating reproducible splits\n",
    "\n",
    "best_acc_val=[]   #checking the accuracy in the validation data\n",
    "best_acc_test=[] #checking the accuracy in the test data\n",
    "best_neighbors=[]    #finding the best k-neighbor value\n",
    "best_precision_test=[] #checking the precision value of test data\n",
    "best_recall_test=[]  #checkig the recall value of the test data\n",
    "best_mcc_test=[]   #checking the mcc value of the test data\n",
    "best_auc_test=[]  #checking Area under the curve (AUC) values of the test data\n",
    "\n",
    "#training data\n",
    "#spliting X_strip and y_strip data for the knn (data after removing the missing values from X and y)\n",
    "for train_index,test_index in kf.split(X_strip,y):\n",
    "    X_strip_train=X_strip[train_index]\n",
    "    X_strip_test=X_strip[test_index]\n",
    "    y_train=y_strip[train_index]\n",
    "    y_test=y_strip[test_index]\n",
    "    \n",
    "    #create train-test split for hyperparameter optimization\n",
    "    X_strip_subtrain,X_strip_val,y_subtrain,y_val=train_test_split(X_strip_train,y_train,test_size=0.2,random_state=42)\n",
    "    neighbors=np.arange(1,1076)   #create a pool for how many neighbors we like to consider\n",
    "    \n",
    "    acc_train=[]      #checking the accuracy of the train data\n",
    "    acc_val=[]   #checking the accuracy of the calidation data\n",
    "    \n",
    "    scores = model.decision_function(X_strip_test)   #scoring for selecting the best model\n",
    "    fpr, tpr, thresholds = metrics.roc_curve(y_test, scores, drop_intermediate=False)\n",
    "\n",
    "\n",
    "    #predicting best k neighbor\n",
    "    for k in neighbors:\n",
    "        knn=KNeighborsClassifier(n_neighbors=k,metric=\"minkowski\",p=2)\n",
    "#metric= minkowski is used because knn use distance to find similarities or dissimilarities of the data and minkowski is the generalization of Euclidean distance\n",
    "#Parameter P value of Minkowski distance set to Euclidian distance\n",
    "\n",
    "        knn.fit(X_strip_subtrain,y_subtrain)     #for training we provide the X_subtrain & y_subtrain, since we haven't optimized the hyperparameter yet              \n",
    "        predictions_training=knn.predict(X_strip_subtrain) #predict for the X_subtrain\n",
    "        predictions_testing=knn.predict(X_strip_val)  #predict for the test value of X_subtrain of this particular split, which is X_val\n",
    "        \n",
    "        #store accuracy\n",
    "        acc_train.append(metrics.accuracy_score(y_subtrain,predictions_training))\n",
    "        acc_val.append(metrics.accuracy_score(y_val,predictions_testing))\n",
    "        \n",
    "    #store information to best model in line-search\n",
    "    best_k=neighbors[np.argmax(acc_val)]\n",
    "    best_neighbors.append(best_k)\n",
    "    best_acc_val.append(acc_val[np.argmax(acc_val)])\n",
    "    \n",
    "    #retrain Model with best k and predict on test data\n",
    "    knn=KNeighborsClassifier(n_neighbors=best_k,metric=\"minkowski\",p=2)\n",
    "    knn.fit(X_strip_train,y_train) #for training we provide the X_strip_train & y_strip_train from the first split done before data reduction step\n",
    "    y_prediction=knn.predict(X_strip_test)\n",
    "    test_accuracy=metrics.accuracy_score(y_test,y_prediction)\n",
    "    #getting all the best scorings for each scoring scheme\n",
    "    best_acc_test.append(test_accuracy)\n",
    "    best_precision_test.append(metrics.precision_score(y_test,y_prediction))\n",
    "    best_recall_test.append(metrics.recall_score(y_test,y_prediction))\n",
    "    best_mcc_test.append(metrics.matthews_corrcoef(y_test,y_prediction))\n",
    "    best_auc_test.append(metrics.roc_auc_score(y_test,y_prediction))\n",
    "    \n",
    "best_neighbors=np.array(best_neighbors)\n",
    "best_acc_val=np.array(best_acc_val)\n",
    "best_acc_test=np.array(best_acc_test)\n",
    "best_precision_test=np.array(best_precision_test)\n",
    "best_recall_test=np.array(best_recall_test)\n",
    "best_auc_test=np.array(best_auc_test)\n",
    "best_mcc_test=np.array(best_mcc_test)\n",
    "        \n",
    "print(\"KNearst Neighbors\")\n",
    "\n",
    "print(\"Average k: %.2f (+- %.2f)\" % (best_neighbors.mean(),best_neighbors.std()))\n",
    "print(\"Average Accuracy (val): %.2f (+- %.2f)\" % (best_acc_val.mean(),best_acc_val.std()))\n",
    "print(\"Average Accuracy (test): %.2f (+- %.2f)\" % (best_acc_test.mean(),best_acc_test.std()))\n",
    "print(\"Average Precision (test): %.2f (+- %.2f)\" % (best_precision_test.mean(),best_precision_test.std()))\n",
    "print(\"Average Recall (test): %.2f (+- %.2f)\" % (best_recall_test.mean(),best_recall_test.std()))\n",
    "print(\"Average ROC-AUC (test): %.2f (+- %.2f)\" % (best_auc_test.mean(),best_auc_test.std()))\n",
    "print(\"Average MCC (test): %.2f (+- %.2f)\" % (best_mcc_test.mean(),best_mcc_test.std()))\n",
    "\n",
    "#generate figure\n",
    "fig = pl.figure(figsize=(7,7)) #axis of the plot\n",
    "ax = fig.add_subplot(111) #adding subplot\n",
    "\n",
    "#plot roc curve for all 10 foldes\n",
    "pl.plot(fpr, tpr, label=\"K-Nearest Neighbor (AUC=%.2f)\" % metrics.roc_auc_score(y_strip_test, y_pred))\n",
    "ax.plot([0,1], [0,1], color=\"grey\",label=\"Random Classifier\",linestyle=\"--\")\n",
    "\n",
    "#Set axis labels\n",
    "ax.set_xlabel(\"False Positive Rate\")\n",
    "ax.set_ylabel(\"True Positive Rate\")\n",
    "#Set axis limits\n",
    "ax.set_ylim(0,1)\n",
    "ax.set_xlim(0,1)\n",
    "#show grid in grey and set top and right axis to invisible\n",
    "ax.grid(color=\"#CCCCCC\")\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "pl.legend()\n",
    "pl.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "445df49e",
   "metadata": {},
   "source": [
    "# Support Vector Machine Using Linear Kernal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "629c8b98",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM Linear Kernel\n",
      "ROC-AUC:\t0.73 (+-0.02)\n",
      "Accuracy:\t0.92 (+-0.01)\n",
      "Recall:\t\t0.47 (+-0.05)\n",
      "Precision:\t0.79 (+-0.04)\n",
      "MCC:\t\t0.57 (+-0.04)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV  #StratifiedKFold is for cross validation by spliting data into train/test\n",
    "from sklearn.svm import SVC #SVC is the support vector machine algorithm for the multiclass problem\n",
    "\n",
    "params = {\"C\": np.logspace(-4,4,10)} #The dictionary has one key \"C\". C can take 10 different values in log-scale\n",
    "\n",
    "#10 fold cross validation by stratifiedKFold\n",
    "cv = StratifiedKFold(n_splits=10,shuffle=True,random_state=42)\n",
    "#n_splits=10 means 10 folds cross validation, shuffle is true means it will move to each of the fold,random_state ensure that the splits are reproducible\n",
    "\n",
    "list_fpr_svm = [] #list for false positive result of svm\n",
    "list_tpr_svm = [] #list for true positive result of svm\n",
    "auc_values_svm = [] #list for area under the curve of svm\n",
    "precision_values = [] \n",
    "recall_values = []\n",
    "accuracy_values = []\n",
    "mcc_values = []\n",
    "\n",
    "####### cross validation by train and validation data spliting X_strip (data after removing missing values of X) and y_strip(data after removing missing values of y)\n",
    "for train_index, val_index in cv.split(X_strip,y):\n",
    "    #Split data\n",
    "    X_strip_train = X_strip[train_index]\n",
    "    X_strip_val = X_strip[val_index]\n",
    "    y_train = y[train_index]\n",
    "    y_val = y[val_index]\n",
    "    \n",
    "    #Train Model by linear kernel\n",
    "    svc = SVC(kernel=\"linear\")\n",
    "    #gridsearch is used for finding out optimal hyperparameters of a model which results in the most accurate prediction\n",
    "    #gridsearchCV through hyperperameter loop fit model on the training set\n",
    "    gridsearch = GridSearchCV(svc,params, cv=2, scoring=\"accuracy\") #cross validation iterators of 2 folds\n",
    "    #the scoring is based on the scoring arguments, such as here \"accuracy\" metrics scorer \n",
    "    gridsearch.fit(X_strip_train,y_train)\n",
    "    \n",
    "    #compute metrics for ROC Curve\n",
    "    scores = gridsearch.decision_function(X_strip_val)\n",
    "    fpr, tpr, thresholds = metrics.roc_curve(y_val, scores, drop_intermediate=False)\n",
    "    \n",
    "    list_fpr_svm.append(fpr)\n",
    "    list_tpr_svm.append(tpr)\n",
    "    \n",
    "    y_pred = gridsearch.predict(X_strip_val)\n",
    "    auc_values_svm.append(metrics.roc_auc_score(y_val, y_pred))\n",
    "    accuracy_values.append(metrics.accuracy_score(y_val, y_pred))\n",
    "    recall_values.append(metrics.recall_score(y_val, y_pred))\n",
    "    precision_values.append(metrics.precision_score(y_val, y_pred))\n",
    "    mcc_values.append(metrics.matthews_corrcoef(y_val, y_pred))\n",
    "\n",
    "\n",
    "#print results\n",
    "print(\"SVM Linear Kernel\")\n",
    "print(\"ROC-AUC:\\t%.2f (+-%.2f)\" % (np.mean(auc_values_svm),np.std(auc_values_svm)))\n",
    "print(\"Accuracy:\\t%.2f (+-%.2f)\" % (np.mean(accuracy_values),np.std(accuracy_values)))\n",
    "print(\"Recall:\\t\\t%.2f (+-%.2f)\" % (np.mean(recall_values),np.std(recall_values)))\n",
    "print(\"Precision:\\t%.2f (+-%.2f)\" % (np.mean(precision_values),np.std(precision_values)))\n",
    "print(\"MCC:\\t\\t%.2f (+-%.2f)\" % (np.mean(mcc_values),np.std(mcc_values)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84fc1ea0",
   "metadata": {},
   "source": [
    "Generate plot for SVM (Linear Kernel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "e0a557e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 700x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = pl.figure(figsize=(7,7)) #represent axis\n",
    "ax = fig.add_subplot(111) #adding subplot\n",
    "\n",
    "#plot roc curve for all 10 foldes\n",
    "pl.plot(fpr, tpr, color=\"#e67e22\",label=\"Mean SVM(Linear Kernel) (10 folds) = %.2f (+- %.2f)\" % (np.mean(auc_values_svm),np.std(auc_values_svm)))\n",
    "ax.plot([0,1], [0,1], color=\"grey\",label=\"Random Classifier\",linestyle=\"--\")\n",
    "\n",
    "\n",
    "#Set axis labels\n",
    "ax.set_xlabel(\"False Positive Rate\")\n",
    "ax.set_ylabel(\"True Positive Rate\")\n",
    "#Set axis limits\n",
    "ax.set_ylim(0,1)\n",
    "ax.set_xlim(0,1)\n",
    "\n",
    "#show grid in grey and set top and right axis to invisible\n",
    "ax.grid(color=\"#CCCCCC\")\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "pl.legend() #add legend\n",
    "pl.tight_layout()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e7cd3ce",
   "metadata": {},
   "source": [
    "# Support vector mechine using RBF-kernel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "ba389b24",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM with a RBF\n",
      "ROC-AUC:\t0.92 (+-0.02)\n",
      "Accuracy:\t0.98 (+-0.01)\n",
      "Recall:\t\t0.84 (+-0.05)\n",
      "Precision:\t0.95 (+-0.02)\n",
      "MCC:\t\t0.88 (+-0.03)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "#Create a dictionary for:-\n",
    "#An array of C values, where C can take 10 different values from log-scale of -4 to 4\n",
    "# addition parameter gamma needed for rbf kernel, which defines how far the single training data can reach\n",
    "params = {\"C\": np.logspace(-4,4,10),\n",
    "          \"gamma\": np.logspace(-4,4,10)} \n",
    "\n",
    "\n",
    "\n",
    "#10 fold cross validation by stratifiedKFold\n",
    "cv = StratifiedKFold(n_splits=10,shuffle=True,random_state=42)\n",
    "#n_splits=10 means 10 folds cross validation, shuffle is true means it will move to each of the fold,random_state ensure that the splits are reproducible\n",
    "\n",
    "\n",
    "list_fpr_svm_rbf = [] #list for false positive result of svm rbf kernel\n",
    "list_tpr_svm_rbf = [] #list for true postitive result of svm rbf kernel\n",
    "auc_values_svm_rbf = [] #list for area under the curve of svm rbf kernel\n",
    "precision_values = [] #list for precision values of svm rbf kernel\n",
    "recall_values = [] #list for recall values of svm rbf kernel\n",
    "accuracy_values = [] #list for accuracy values of svm rbf kernel\n",
    "mcc_values = [] #list for mcc values of svm rbf kernel\n",
    "\n",
    "####### cross validation by train and validation data spliting X_strip and y_strip\n",
    "for train_index, val_index in cv.split(X_strip,y):\n",
    "    #Split data\n",
    "    X_strip_train = X_strip[train_index]\n",
    "    X_strip_val = X_strip[val_index]\n",
    "    y_train = y[train_index]\n",
    "    y_val = y[val_index]\n",
    "    \n",
    "    #Train Model by rbf kernel\n",
    "    svc = SVC(kernel=\"rbf\")\n",
    "    #gridsearch is used for finding out optimal hyperparameters of a model which results in the most accurate prediction\n",
    "    #gridsearchCV through hyperperameter loop fit model on the training set\n",
    "    gridsearch = GridSearchCV(svc,params, cv=2, scoring=\"accuracy\")  #cross validation iterators of 2 folds\n",
    "    #the scoring is based on the scoring arguments, such as here \"accuracy\" metrics scorer \n",
    "    gridsearch.fit(X_strip_train,y_train)\n",
    "    \n",
    "    #compute metrics for ROC Curve\n",
    "    scores = gridsearch.decision_function(X_strip_val)\n",
    "    fpr, tpr, thresholds = metrics.roc_curve(y_val, scores, drop_intermediate=False)\n",
    "    \n",
    "    list_fpr_svm_rbf.append(fpr)\n",
    "    list_tpr_svm_rbf.append(tpr)\n",
    "    \n",
    "    y_pred = gridsearch.predict(X_strip_val)\n",
    "    auc_values_svm_rbf.append(metrics.roc_auc_score(y_val, y_pred))\n",
    "    accuracy_values.append(metrics.accuracy_score(y_val, y_pred))\n",
    "    recall_values.append(metrics.recall_score(y_val, y_pred))\n",
    "    precision_values.append(metrics.precision_score(y_val, y_pred))\n",
    "    mcc_values.append(metrics.matthews_corrcoef(y_val, y_pred))\n",
    "\n",
    "\n",
    "#print results\n",
    "print(\"SVM with a RBF\")\n",
    "print(\"ROC-AUC:\\t%.2f (+-%.2f)\" % (np.mean(auc_values_svm_rbf),np.std(auc_values_svm_rbf)))\n",
    "print(\"Accuracy:\\t%.2f (+-%.2f)\" % (np.mean(accuracy_values),np.std(accuracy_values)))\n",
    "print(\"Recall:\\t\\t%.2f (+-%.2f)\" % (np.mean(recall_values),np.std(recall_values)))\n",
    "print(\"Precision:\\t%.2f (+-%.2f)\" % (np.mean(precision_values),np.std(precision_values)))\n",
    "print(\"MCC:\\t\\t%.2f (+-%.2f)\" % (np.mean(mcc_values),np.std(mcc_values)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eac33301",
   "metadata": {},
   "source": [
    "Generate plot for SVM (RBF Kernel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "3689b1f4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 700x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = pl.figure(figsize=(7,7))  #represent axis\n",
    "ax = fig.add_subplot(111) #adding subplot\n",
    "\n",
    "#plot roc curve for all 10 foldes\n",
    "pl.plot(fpr, tpr, color=\"#e67e22\",label=\"Mean SVM(RBF Kernel) (10 folds) = %.2f (+- %.2f)\" % (np.mean(auc_values_svm_rbf),np.std(auc_values_svm_rbf)))\n",
    "ax.plot([0,1], [0,1], color=\"grey\",label=\"Random Classifier\",linestyle=\"--\")\n",
    "\n",
    "\n",
    "#Set axis labels\n",
    "ax.set_xlabel(\"False Positive Rate\")\n",
    "ax.set_ylabel(\"True Positive Rate\")\n",
    "#Set axis limits\n",
    "ax.set_ylim(0,1)\n",
    "ax.set_xlim(0,1)\n",
    "\n",
    "#show grid in grey and set top and right axis to invisible\n",
    "ax.grid(color=\"#CCCCCC\")\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "pl.legend()\n",
    "pl.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c236497d",
   "metadata": {},
   "source": [
    "# Importing Unknown data and Prediction of Toxicity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "6662d6c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "unknown = np.array(np.loadtxt(\"unknown_data.csv\", delimiter=\",\", dtype = np.str)) #imporing unknown_data.csv into array\n",
    "unknown_sample=unknown[1:,0] #sample elements(names) of the unknown data\n",
    "unknown=np.float_(unknown[1:,1:]) #converting unknown string data (excluding the feature and label names) into float data \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7adab278",
   "metadata": {},
   "source": [
    "Checking Unknown data ( Sample number, feature number and missing values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "74193ee5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples of Unknown data:610\t\t\n",
      "Number of features of unknown data:801\t\t\n",
      "Missing values of unknown: 0 \t\t\n",
      "Percentage of Missing values of unknown: 0\t\t\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of samples of Unknown data:%d\\t\\t\"%unknown.shape[0]) #unknown samples\n",
    "print(\"Number of features of unknown data:%d\\t\\t\"%unknown.shape[1]) #unknown features\n",
    "missing_values=(np.isnan(unknown)).sum() #isnan() for checking missing values of unknown data\n",
    "print(\"Missing values of unknown: %d \\t\\t\"%missing_values)\n",
    "percentage_missing=(missing_values/unknown.size)*100\n",
    "print(\"Percentage of Missing values of unknown: %d\\t\\t\"%percentage_missing)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "89a7cfe1",
   "metadata": {},
   "source": [
    "### Normalizing Unknown data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "810cd012",
   "metadata": {},
   "outputs": [],
   "source": [
    "unknown = scaler.fit_transform(unknown)  #fit is to calculating the mean and varience of each features and transform function returns the dataset produced by it\n",
    "                                      #fit_transform is used together to improve the efficiency of the model\n",
    "                                #scaler for data normalization"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9eaf945",
   "metadata": {},
   "source": [
    " ### Predicting y label for Unknown data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5c25d0cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_unknown = gridsearch.predict(unknown)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "181e3dde",
   "metadata": {},
   "source": [
    "### Generating CSV file"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65e2c194",
   "metadata": {},
   "source": [
    "Generating definition output for CSV file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f5a17980",
   "metadata": {},
   "outputs": [],
   "source": [
    "i = 0\n",
    "output = \"\"\n",
    "for val in unknown_sample:\n",
    "    output = output + str(unknown_sample[i]) + \",\" +str(int(y_unknown[i])) + \"\\n\"\n",
    "    i += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "88df2781",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "try: os.remove(\"Unkwon_RBF.csv\") #removes file Unkwon_RBF.csv, if it exists\n",
    "except: pass # passes if the file doesn't exist\n",
    "open(\"Unkwon_RBF_88 .csv\", 'a').write(str(output)) #opens the file & writes the definition output into it"
   ]
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}