Maitreyee-FinalSolution-Multiclass-Prediction-Model-LASI-DAD.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Neurohack - Challenge 2 London team E\n",
    "### Author: Maitreyee Wairagkar\n",
    "Last update: 14/01/2022\n",
    "\n",
    "## PCA + Multiclass (5-class) SVM based Clinical Dementia Rating Prediction Model - Complete Pipeline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from mpl_toolkits import mplot3d\n",
    "\n",
    "'''\n",
    "# Convert .dat to csv\n",
    "data = pd.io.stata.read_stata('H_DAD_w1a3.dta')\n",
    "data.to_csv('H_DAD_w1a3.csv')\n",
    "'''\n",
    "\n",
    "data = pd.read_csv('H_DAD_w1a3.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Pre-processing by removing non-relevant features and NaN columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Features to reject\n",
    "remove_features = ['prim_key','hhid','pnc','pn','r1stateid','r1iwy_d','r1iwm_d','r1lasidy','rabyear', \n",
    "                   'rabmonth', 'r1lang_d', 'r1location', 'r1mheight', 'r1mweight', 'r1ht_flag', \n",
    "                   'r1wt_flag', 'r1htlimbs', 'r1midarm', 'r1calf', 'r1kneeht', 'r1raterid1', \n",
    "                   'r1raterid2', 'r1raterid3', 'r1phase', 'r1iwstat_d', 'r1inf_age', 'r1inf_gendr', \n",
    "                   'r1inf_educ', 'r1inf_rel', 'r1inf_freq', 'r1inf_care', 'r1inf_yrs', 'r1mbmi', 'r1bmicat',] \n",
    "\n",
    "potential_labels = ['r1cdr_final','r1cdr_incon','r1cdr_mem1','r1cdr_ori1','r1cdr_jud1','r1cdr_com1',\n",
    "                    'r1cdr_hom1','r1cdr_per1','r1cdr_scor1','r1cdr_mem2','r1cdr_ori2','r1cdr_jud2',\n",
    "                    'r1cdr_com2','r1cdr_hom2','r1cdr_per2','r1cdr_scor2','r1cdr_mem3','r1cdr_ori3',\n",
    "                    'r1cdr_jud3','r1cdr_com3','r1cdr_hom3','r1cdr_per3','r1cdr_scor3']\n",
    "\n",
    "\n",
    "# Important features to select\n",
    "features_to_select=['r1agey','ragender','raedyrs','h1rural','r1location','r1wtresp','r1obsnoise','r1obsodor',\n",
    "                    'r1obsair','r1obshouse','r1borient','r1bexefu','r1blangf','r1bmemory','r1bvsp','r1nmemimm',\n",
    "                    'r1nmemdel','r1nmemrec','r1nreason','r1natnspd','r1sgcp','r1hmse_scorz','r1word_totaz',\n",
    "                    'r1word_dz','r1wre_scorez','r1log_recoz','r1bm_immexz','r1bm_reclexz','r1verbalz',\n",
    "                    'r1csid_scorz','r1rv_scorez','r1cog_totalz','r1i_hear','r1i_sleep','r1systo','r1diasto',\n",
    "                    'r1pulse','r1bphigh','r1mheight','r1mweight','r1mbmi','r1bmicat','r1adla_d','r1iadltot1_d',\n",
    "                    'r1cesd10','r1anx5','r1mna_scale','r1spice','r1hear_r','r1hear_l','r1hear_aid','r1prs_toplam',\n",
    "                    'r1prs_topkun','r1prs_topjan','r1prs_topcog','r1prs_pc1','r1prs_pc2','r1prs_pc3','r1prs_pc4',\n",
    "                    'r1prs_pc5','r1rs7412','r1rs429358']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2528, 490)"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pre-processing: remove non-important features\n",
    "\n",
    "# remove rows with missing values for clinician evaluation (this discards phase 1)\n",
    "df = data.dropna(subset = ['r1cdr_final'])\n",
    "\n",
    "labels = df['r1cdr_final']\n",
    "gender = df['ragender']\n",
    "\n",
    "df = df.drop(columns=potential_labels)\n",
    "df = df.drop(columns=remove_features)\n",
    "\n",
    "# remove columns with NaN\n",
    "df = df.dropna(axis=1)\n",
    "\n",
    "'''\n",
    "# keep only selected features \n",
    "df = df.filter(items=features_to_select)\n",
    "# Note that most of the above columns have been eliminated previously because of NaNs\n",
    "df.shape\n",
    "'''\n",
    "\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.1 Data Visualisation: T-SNE Visualisation of Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compute and plot T-SNE\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "# We want to get TSNE embedding with 2 dimensions\n",
    "n_components = 3\n",
    "tsne = TSNE(n_components)\n",
    "tsne_result = tsne.fit_transform(df)\n",
    "tsne_result.shape\n",
    "\n",
    "# Plot T-SNE\n",
    "scatter_x = np.array(tsne_result[:,0])\n",
    "scatter_y = np.array(tsne_result[:,1])\n",
    "scatter_z = np.array(tsne_result[:,2])\n",
    "group = np.array(labels)\n",
    "cdict = {0: 'green', 0.5: 'blue', 1:'orange', 2:'red', 3: 'purple'}\n",
    "\n",
    "fig = plt.figure(figsize = (7, 7))\n",
    "ax = plt.axes(projection =\"3d\")\n",
    "\n",
    "for g in np.unique(group):\n",
    "    ix = np.where(group == g)\n",
    "    ax.scatter3D(scatter_x[ix], scatter_y[ix], scatter_z[ix], c = cdict[g], label = g)\n",
    "ax.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.2 Data Visualisation: Histogram of pre-processed dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram\n",
    "# Creating dataset\n",
    "a = []\n",
    "for i in labels:\n",
    "    if i == 0: # no dementia\n",
    "        a.append(0)\n",
    "    if i == 0.5: # questionable\n",
    "        a.append(1)\n",
    "    if i == 1: # mild\n",
    "        a.append(2)\n",
    "    if i == 2: # moderate\n",
    "        a.append(3)\n",
    "    if i == 3: # severe\n",
    "        a.append(4)\n",
    "\n",
    "# Creating histogram\n",
    "labl, counts = np.unique(a, return_counts=True)\n",
    "\n",
    "fig = plt.figure(figsize = (9, 6))\n",
    "plt.rc('font', size=15) \n",
    "\n",
    "plt.bar(labl, counts, align='center')\n",
    "plt.gca().set_xticks(labl)\n",
    "plt.gca().set_xticklabels(('No\\nimpairment', 'Questionable\\nimpairment', 'Mild\\ndementia', 'Moderate\\ndementia', 'Severe\\ndementia'))\n",
    "plt.xlabel('\\nConsensus Clinical Dementia Rating Score')\n",
    "plt.ylabel('Number of people')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.3 Data Visualisation: PCA on cleaned data (all features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f870431e9d0>]"
      ]
     },
     "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": [
    "# standardise the data before PCA\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "x = MinMaxScaler().fit_transform(df) # normalizing the features\n",
    "x.shape\n",
    "\n",
    "\n",
    "# PCA on all features\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "pca = PCA() #n_components=3\n",
    "principalComponents = pca.fit_transform(x)\n",
    "\n",
    "plt.plot(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot PCA components\n",
    "\n",
    "#%matplotlib qt\n",
    "\n",
    "comp =[0,1,2] # select components to plot\n",
    "scatter_x = np.array(principalComponents[:,comp[0]])\n",
    "scatter_y = np.array(principalComponents[:,comp[1]])\n",
    "scatter_z = np.array(principalComponents[:,comp[2]])\n",
    "group = np.array(labels)\n",
    "cdict = {0: 'green', 0.5: 'blue', 1:'orange', 2:'red', 3: 'purple'}\n",
    "label_names = {0:'No impairment', 0.5:'Questionable impairment', 1:'Mild dementia', 2:'Moderate dementia', 3:'Severe dementia'}\n",
    "\n",
    "fig = plt.figure(figsize = (9, 9))\n",
    "plt.rc('font', size=15) \n",
    "ax = plt.axes(projection =\"3d\")\n",
    "\n",
    "\n",
    "for g in np.unique(group):\n",
    "    ix = np.where(group == g)\n",
    "    ax.scatter(scatter_x[ix], scatter_y[ix], scatter_z[ix], c = cdict[g], label = label_names[g] )\n",
    "    \n",
    "ax.legend(bbox_to_anchor=(0, 1.15),loc='upper left', title='Clinical Dementia Rating')\n",
    "plt.xlabel('Principal Component '+str(comp[0]+1))\n",
    "plt.ylabel('Principal Component '+str(comp[1]+1))\n",
    "ax.set_zlabel('Principal Component '+str(comp[2]+1))\n",
    "\n",
    "ax.view_init(0, 69)\n",
    "\n",
    "ax.w_xaxis.set_ticklabels([])\n",
    "ax.w_yaxis.set_ticklabels([])\n",
    "ax.w_zaxis.set_ticklabels([])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Write PCA output in csv\n",
    "pc_d=pd.DataFrame(principalComponents)\n",
    "pc_d.to_csv('principalComponents.csv')\n",
    "\n",
    "l=pd.DataFrame({'CDR_score':labels,\n",
    "                    'gender': gender})\n",
    "l.to_csv('data_labels.csv')\n",
    "\n",
    "p=pd.DataFrame(pca.components_)\n",
    "p.to_csv('PCAweights.csv')\n",
    "\n",
    "v=pd.DataFrame(pca.explained_variance_ratio_)\n",
    "v.to_csv('PCAvariance.csv')\n",
    "\n",
    "col = pd.DataFrame(df.columns)\n",
    "col.to_csv('PCAcolumns.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Multiclass classification using PCA+SVM with 10-fold cross-validation\n",
    "## 4.1 Input Data Preparation for Multiclass Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7869, 491)\n"
     ]
    },
    {
     "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": [
    "import numpy.matlib\n",
    "from sklearn.model_selection import train_test_split, KFold, StratifiedKFold, cross_val_score\n",
    "from sklearn import svm\n",
    "from sklearn import metrics\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import confusion_matrix,classification_report\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "# 1. prepare data \n",
    "group = np.array(labels)\n",
    "\n",
    "# get indices for different classes\n",
    "for g in np.unique(group):\n",
    "    ix = np.where(group == g)\n",
    "    if g == 0:\n",
    "        class_1 = ix\n",
    "    if g == 0.5: \n",
    "        class_2 = ix\n",
    "    if g == 1:\n",
    "        class_3 = ix\n",
    "    if g == 2: \n",
    "        class_4 = ix\n",
    "    if g == 3:\n",
    "        class_5 = ix\n",
    "        \n",
    "dat = df.to_numpy()\n",
    "        \n",
    "# Get data for each class and add class label in the last column #principalComponents\n",
    "dat_1 = np.concatenate((np.squeeze(dat[class_1,:]),np.ones([len(np.squeeze(class_1)),1])),axis=1)\n",
    "dat_2 = np.concatenate((np.squeeze(dat[class_2,:]),np.ones([len(np.squeeze(class_2)),1])*2),axis=1)\n",
    "dat_3 = np.concatenate((np.squeeze(dat[class_3,:]),np.ones([len(np.squeeze(class_3)),1])*3),axis=1)\n",
    "dat_4 = np.concatenate((np.squeeze(dat[class_4,:]),np.ones([len(np.squeeze(class_4)),1])*4),axis=1)\n",
    "dat_5 = np.concatenate((np.squeeze(dat[class_5,:]),np.ones([len(np.squeeze(class_5)),1])*5),axis=1)\n",
    "\n",
    "# up sampling: repeat matrices to match number of samples in largest class\n",
    "dat1 = np.matlib.repmat(dat_1,int(np.round(dat_2.shape[0]/dat_1.shape[0])),1)\n",
    "dat2 = dat_2\n",
    "dat3 = np.matlib.repmat(dat_3,int(np.round(dat_2.shape[0]/dat_3.shape[0])),1)\n",
    "dat4 = np.matlib.repmat(dat_4,int(np.round(dat_2.shape[0]/dat_4.shape[0])),1)\n",
    "dat5 = np.matlib.repmat(dat_5,int(np.round(dat_2.shape[0]/dat_5.shape[0])),1)\n",
    "\n",
    "# collect dat from all classes together and shuffle the order\n",
    "all_dat = []\n",
    "all_dat = np.concatenate((dat1,dat2,dat3,dat4,dat5),axis=0) \n",
    "np.random.shuffle(all_dat)\n",
    "plt.hist(all_dat[:,-1])\n",
    "print(all_dat.shape)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.2. Hyperparameter tuning using Grid Search for multiclass SVM model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Hyperparameter tuning for SVM model\n",
    "\n",
    "from sklearn.model_selection import cross_val_score, GridSearchCV\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "# Create the parameter grid based on the results of random search \n",
    "params_grid = [{'kernel': ['rbf'], 'gamma': [1e-3, 1e-4],'C': [0.5, 1, 10, 100, 1000]},\n",
    "               {'kernel': ['linear'], 'C': [0.5, 1, 10, 100, 1000]}]\n",
    "\n",
    "# Performing CV to tune parameters for best SVM fit \n",
    "svm_model = GridSearchCV(SVC(), params_grid, cv=5)\n",
    "\n",
    "X_skf = all_dat[:,0:-1] # data\n",
    "Y_skf = all_dat[:,-1]   # labels\n",
    "\n",
    "train = MinMaxScaler().fit_transform(X_skf)\n",
    "\n",
    "svm_model.fit(train, Y_skf)\n",
    "\n",
    "print(svm_model.best_estimator_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.3 Multiclass classification using PCA+ SVM (all features) 10-fold cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "------------------------------------------------------------------------\n",
      "Training for fold 1 for SVM ...\n",
      "[[146.  13.   0.   0.   0.]\n",
      " [ 16. 130.   3.   0.   0.]\n",
      " [  0.   0. 154.   0.   0.]\n",
      " [  0.   0.   0. 157.   0.]\n",
      " [  0.   0.   0.   0. 168.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.90      0.92      0.91       159\n",
      "         2.0       0.91      0.87      0.89       149\n",
      "         3.0       0.98      1.00      0.99       154\n",
      "         4.0       1.00      1.00      1.00       157\n",
      "         5.0       1.00      1.00      1.00       168\n",
      "\n",
      "    accuracy                           0.96       787\n",
      "   macro avg       0.96      0.96      0.96       787\n",
      "weighted avg       0.96      0.96      0.96       787\n",
      "\n",
      "Training set score for SVM: 0.989692\n",
      "Testing  set score for SVM: 0.959339\n",
      "Loss per fold: 0.040661\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 2 for SVM ...\n",
      "[[277.  25.   0.   0.   0.]\n",
      " [ 22. 254.   5.   0.   0.]\n",
      " [  0.   0. 327.   0.   0.]\n",
      " [  0.   0.   0. 338.   0.]\n",
      " [  0.   0.   0.   0. 326.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.96      0.92      0.94       143\n",
      "         2.0       0.91      0.94      0.93       132\n",
      "         3.0       0.99      1.00      0.99       173\n",
      "         4.0       1.00      1.00      1.00       181\n",
      "         5.0       1.00      1.00      1.00       158\n",
      "\n",
      "    accuracy                           0.97       787\n",
      "   macro avg       0.97      0.97      0.97       787\n",
      "weighted avg       0.97      0.97      0.97       787\n",
      "\n",
      "Training set score for SVM: 0.986586\n",
      "Testing  set score for SVM: 0.974587\n",
      "Loss per fold: 0.025413\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 3 for SVM ...\n",
      "[[420.  35.   0.   0.   0.]\n",
      " [ 38. 391.   6.   0.   0.]\n",
      " [  0.   0. 485.   0.   0.]\n",
      " [  0.   0.   0. 495.   0.]\n",
      " [  0.   0.   0.   0. 491.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.90      0.93      0.92       153\n",
      "         2.0       0.93      0.89      0.91       154\n",
      "         3.0       0.99      1.00      1.00       158\n",
      "         4.0       1.00      1.00      1.00       157\n",
      "         5.0       1.00      1.00      1.00       165\n",
      "\n",
      "    accuracy                           0.97       787\n",
      "   macro avg       0.97      0.96      0.96       787\n",
      "weighted avg       0.97      0.97      0.97       787\n",
      "\n",
      "Training set score for SVM: 0.988563\n",
      "Testing  set score for SVM: 0.965693\n",
      "Loss per fold: 0.034307\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 4 for SVM ...\n",
      "[[535.  56.   0.   0.   0.]\n",
      " [ 61. 553.   7.   0.   0.]\n",
      " [  0.   0. 635.   0.   0.]\n",
      " [  0.   0.   0. 659.   0.]\n",
      " [  0.   0.   0.   0. 642.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.83      0.85      0.84       136\n",
      "         2.0       0.89      0.87      0.88       186\n",
      "         3.0       0.99      1.00      1.00       150\n",
      "         4.0       1.00      1.00      1.00       164\n",
      "         5.0       1.00      1.00      1.00       151\n",
      "\n",
      "    accuracy                           0.94       787\n",
      "   macro avg       0.94      0.94      0.94       787\n",
      "weighted avg       0.94      0.94      0.94       787\n",
      "\n",
      "Training set score for SVM: 0.987151\n",
      "Testing  set score for SVM: 0.942821\n",
      "Loss per fold: 0.057179\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 5 for SVM ...\n",
      "[[676.  73.   0.   0.   0.]\n",
      " [ 73. 704.   9.   0.   0.]\n",
      " [  0.   0. 797.   0.   0.]\n",
      " [  0.   0.   0. 810.   0.]\n",
      " [  0.   0.   0.   0. 793.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.92      0.89      0.91       158\n",
      "         2.0       0.90      0.92      0.91       165\n",
      "         3.0       0.99      1.00      0.99       162\n",
      "         4.0       1.00      1.00      1.00       151\n",
      "         5.0       1.00      1.00      1.00       151\n",
      "\n",
      "    accuracy                           0.96       787\n",
      "   macro avg       0.96      0.96      0.96       787\n",
      "weighted avg       0.96      0.96      0.96       787\n",
      "\n",
      "Training set score for SVM: 0.986303\n",
      "Testing  set score for SVM: 0.960610\n",
      "Loss per fold: 0.039390\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 6 for SVM ...\n",
      "[[802.  88.   0.   0.   0.]\n",
      " [ 90. 840.  11.   0.   0.]\n",
      " [  0.   0. 972.   0.   0.]\n",
      " [  0.   0.   0. 965.   0.]\n",
      " [  0.   0.   0.   0. 954.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.88      0.89      0.89       141\n",
      "         2.0       0.90      0.88      0.89       155\n",
      "         3.0       0.99      1.00      0.99       175\n",
      "         4.0       1.00      1.00      1.00       155\n",
      "         5.0       1.00      1.00      1.00       161\n",
      "\n",
      "    accuracy                           0.96       787\n",
      "   macro avg       0.95      0.95      0.95       787\n",
      "weighted avg       0.96      0.96      0.96       787\n",
      "\n",
      "Training set score for SVM: 0.988280\n",
      "Testing  set score for SVM: 0.956798\n",
      "Loss per fold: 0.043202\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 7 for SVM ...\n",
      "[[ 941.  108.    0.    0.    0.]\n",
      " [ 108.  983.   12.    0.    0.]\n",
      " [   0.    0. 1119.    0.    0.]\n",
      " [   0.    0.    0. 1111.    0.]\n",
      " [   0.    0.    0.    0. 1127.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.89      0.87      0.88       159\n",
      "         2.0       0.88      0.88      0.88       162\n",
      "         3.0       0.99      1.00      1.00       147\n",
      "         4.0       1.00      1.00      1.00       146\n",
      "         5.0       1.00      1.00      1.00       173\n",
      "\n",
      "    accuracy                           0.95       787\n",
      "   macro avg       0.95      0.95      0.95       787\n",
      "weighted avg       0.95      0.95      0.95       787\n",
      "\n",
      "Training set score for SVM: 0.986727\n",
      "Testing  set score for SVM: 0.950445\n",
      "Loss per fold: 0.049555\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 8 for SVM ...\n",
      "[[1097.  130.    0.    0.    0.]\n",
      " [ 119. 1109.   14.    0.    0.]\n",
      " [   0.    0. 1291.    0.    0.]\n",
      " [   0.    0.    0. 1266.    0.]\n",
      " [   0.    0.    0.    0. 1270.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.93      0.88      0.90       178\n",
      "         2.0       0.85      0.91      0.88       139\n",
      "         3.0       0.99      1.00      0.99       172\n",
      "         4.0       1.00      1.00      1.00       155\n",
      "         5.0       1.00      1.00      1.00       143\n",
      "\n",
      "    accuracy                           0.96       787\n",
      "   macro avg       0.95      0.96      0.96       787\n",
      "weighted avg       0.96      0.96      0.96       787\n",
      "\n",
      "Training set score for SVM: 0.986586\n",
      "Testing  set score for SVM: 0.955527\n",
      "Loss per fold: 0.044473\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 9 for SVM ...\n",
      "[[1229.  145.    0.    0.    0.]\n",
      " [ 132. 1256.   15.    0.    0.]\n",
      " [   0.    0. 1456.    0.    0.]\n",
      " [   0.    0.    0. 1429.    0.]\n",
      " [   0.    0.    0.    0. 1421.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.91      0.90      0.90       147\n",
      "         2.0       0.91      0.91      0.91       161\n",
      "         3.0       0.99      1.00      1.00       165\n",
      "         4.0       1.00      1.00      1.00       163\n",
      "         5.0       1.00      1.00      1.00       151\n",
      "\n",
      "    accuracy                           0.96       787\n",
      "   macro avg       0.96      0.96      0.96       787\n",
      "weighted avg       0.96      0.96      0.96       787\n",
      "\n",
      "Training set score for SVM: 0.988704\n",
      "Testing  set score for SVM: 0.963151\n",
      "Loss per fold: 0.036849\n",
      "------------------------------------------------------------------------\n",
      "Training for fold 10 for SVM ...\n",
      "[[1372.  164.    0.    0.    0.]\n",
      " [ 160. 1392.   16.    0.    0.]\n",
      " [   0.    0. 1620.    0.    0.]\n",
      " [   0.    0.    0. 1575.    0.]\n",
      " [   0.    0.    0.    0. 1570.]]\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         1.0       0.84      0.88      0.86       162\n",
      "         2.0       0.88      0.82      0.85       165\n",
      "         3.0       0.99      1.00      1.00       164\n",
      "         4.0       1.00      1.00      1.00       146\n",
      "         5.0       1.00      1.00      1.00       149\n",
      "\n",
      "    accuracy                           0.94       786\n",
      "   macro avg       0.94      0.94      0.94       786\n",
      "weighted avg       0.94      0.94      0.94       786\n",
      "\n",
      "Training set score for SVM: 0.987717\n",
      "Testing  set score for SVM: 0.938931\n",
      "Loss per fold: 0.061069\n"
     ]
    }
   ],
   "source": [
    "# Classification with SVM\n",
    "\n",
    "# Get splits for Kfold cross-validation\n",
    "kf = KFold(n_splits = 10, shuffle = True)\n",
    "\n",
    "# Data\n",
    "X_skf = all_dat[:,0:-1] # data\n",
    "Y_skf = all_dat[:,-1]   # labels\n",
    "\n",
    "# Define the models\n",
    "clf = svm.SVC(kernel='rbf', C=0.5, class_weight='balanced') # C is regularisation\n",
    "pca = PCA() \n",
    "lda = LinearDiscriminantAnalysis()\n",
    "scaler = MinMaxScaler()\n",
    "\n",
    "# Cross-validation and training\n",
    "\n",
    "fold_no = 0\n",
    "cnt = 0\n",
    "cm =np.zeros([5,5]) # initialise empty confusion matrix\n",
    "classi_report = []\n",
    "train_score = []\n",
    "test_score = []\n",
    "loss = []\n",
    "all_y_true = []\n",
    "all_y_pred = []\n",
    "\n",
    "for train_idx, test_idx in kf.split(X_skf, Y_skf): \n",
    "    \n",
    "    print('------------------------------------------------------------------------')\n",
    "    print(f'Training for fold {fold_no+1} for SVM ...')\n",
    "    \n",
    "    # 0. Scale training data\n",
    "    scaler.fit(X_skf[train_idx,:])\n",
    "    dat_train = scaler.transform(X_skf[train_idx,:])\n",
    "    dat_test = scaler.transform(X_skf[test_idx,:])\n",
    "    \n",
    "    # 1. Fit PCA to to train data\n",
    "    pca.fit(dat_train)\n",
    "    pc_train = pca.transform(dat_train)\n",
    "    pc_test = pca.transform(dat_test)\n",
    "    \n",
    "    # 1.1 Scale pcs\n",
    "    scaler.fit(pc_train)\n",
    "    pc_train = scaler.transform(pc_train)\n",
    "    pc_test = scaler.transform(pc_test)\n",
    "    \n",
    "    '''\n",
    "    # 2. Fit LDA to principal components (train data) to enhance separability of features\n",
    "    lda.fit(pc_train, Y_skf[train_idx])\n",
    "    ld_train = lda.transform(pc_train)\n",
    "    ld_test = lda.transform(pc_test)\n",
    "\n",
    "    # 2.2 Scale ld components\n",
    "    scaler.fit(ld_train)\n",
    "    ld_train = scaler.transform(ld_train)\n",
    "    ld_test = scaler.transform(ld_test)\n",
    "    '''\n",
    "    train = pc_train\n",
    "    test = pc_test\n",
    "    \n",
    "    # 3. Fit SVM to LDA fitted components\n",
    "    clf.fit(train, Y_skf[train_idx])\n",
    "    \n",
    "    # 4. Make Prediction with SVM on test data\n",
    "    y_pred = clf.predict(test)\n",
    "    \n",
    "    # 6. Evaluate the performance of SVM with Confusion Matrix and record performance\n",
    "    cm = cm + confusion_matrix(Y_skf[test_idx],y_pred)\n",
    "    classi_report.append(classification_report(Y_skf[test_idx],y_pred))\n",
    " \n",
    "    train_score.append(clf.score(train, Y_skf[train_idx]))\n",
    "    test_score.append(clf.score(test, Y_skf[test_idx]))\n",
    "    loss.append(metrics.mean_squared_error(Y_skf[test_idx], y_pred))\n",
    "    \n",
    "    all_y_true.append(Y_skf[test_idx])\n",
    "    all_y_pred.append(y_pred)\n",
    "    \n",
    "    print(cm)\n",
    "    print('\\n')\n",
    "    print(classi_report[cnt])\n",
    "    \n",
    "    print(\"Training set score for SVM: %f\" %train_score[cnt] )\n",
    "    print(\"Testing  set score for SVM: %f\" %test_score[cnt] )\n",
    "    print(\"Loss per fold: %f\" %loss[cnt])\n",
    "    cnt=cnt+1\n",
    "  \n",
    "    fold_no+=1\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.4 Print Classification Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix for multiclass classifier (%):\n",
      "\n",
      "                No dementia  Questionable  Mild   Moderate  Severe  \n",
      "No dementia     89.32        10.68          0.0    0.0       0.0\n",
      "Questionable    10.2        88.78         1.02    0.0       0.0\n",
      "Mild            0.0          0.0          100.0  0.0       0.0\n",
      "Moderate        0.0          0.0           0.0    100.0     0.0\n",
      "Severe.         0.0          0.0           0.0    0.0       100.0\n",
      "--------------------------------------------------------------------\n",
      "\n",
      "Model Performance: \n",
      "\n",
      "Class                       Precision       Recall      f1-score\n",
      "No dementia                 89.56           89.32       89.44\n",
      "Questionable impairment     89.46           88.78       89.12\n",
      "Mild dementia               99.02           100.0       99.51\n",
      "Moderate dementia           100.0           100.0       100.0\n",
      "Severe dementia             100.0           100.0       100.0\n",
      "\n",
      "Accuracy on test data: 95.68%\n",
      "Accuracy on training data: 98.76%\n",
      "\n",
      "MSE Loss: 0.043\n"
     ]
    }
   ],
   "source": [
    "# Print results \n",
    "\n",
    "CM = confusion_matrix(np.hstack(all_y_true),np.hstack(all_y_pred),normalize='true')*100\n",
    "CM = np.round(CM,2)\n",
    "print(\"Confusion matrix for multiclass classifier (%):\\n\")\n",
    "print(\"                No dementia  \"+ 'Questionable  '+ 'Mild   '+ 'Moderate  '+ 'Severe  ')\n",
    "print(\"No dementia     \"+ str(CM[0,0]) + \"        \" + str(CM[0,1])+ \"          \" + str(CM[0,2])+ \"    \" + str(CM[0,3])+ \"       \" + str(CM[0,4]))\n",
    "print(\"Questionable    \"+ str(CM[1,0]) + \"        \" + str(CM[1,1])+ \"         \" + str(CM[1,2])+ \"    \" + str(CM[1,3])+ \"       \" + str(CM[1,4]))\n",
    "print(\"Mild            \"+ str(CM[2,0]) + \"          \" + str(CM[2,1])+ \"          \" + str(CM[2,2])+ \"  \" + str(CM[2,3])+ \"       \" + str(CM[2,4]))\n",
    "print(\"Moderate        \"+ str(CM[3,0]) + \"          \" + str(CM[3,1])+ \"           \" + str(CM[3,2])+ \"    \" + str(CM[3,3])+ \"     \" + str(CM[3,4]))\n",
    "print(\"Severe.         \"+ str(CM[4,0]) + \"          \" + str(CM[4,1])+ \"           \" + str(CM[4,2])+ \"    \" + str(CM[4,3])+ \"       \" + str(CM[4,4]))\n",
    "\n",
    "print('--------------------------------------------------------------------')\n",
    "print()\n",
    "print('Model Performance: ')\n",
    "print()\n",
    "report = classification_report(np.hstack(all_y_true),np.hstack(all_y_pred), output_dict=True)\n",
    "print(\"Class                       Precision       Recall      f1-score\")\n",
    "print(\"No dementia                 \"+ str(np.round(report['1.0']['precision']*100,2))+\"           \"+\n",
    "                                      str(np.round(report['1.0']['recall']*100,2)) +\"       \"+\n",
    "                                      str(np.round(report['1.0']['f1-score']*100,2)))\n",
    "print(\"Questionable impairment     \"+ str(np.round(report['2.0']['precision']*100,2))+\"           \"+\n",
    "                                      str(np.round(report['2.0']['recall']*100,2)) +\"       \"+\n",
    "                                      str(np.round(report['2.0']['f1-score']*100,2)))\n",
    "print(\"Mild dementia               \"+ str(np.round(report['3.0']['precision']*100,2))+\"           \"+\n",
    "                                      str(np.round(report['3.0']['recall']*100,2)) +\"       \"+\n",
    "                                      str(np.round(report['3.0']['f1-score']*100,2)))\n",
    "print(\"Moderate dementia           \"+ str(np.round(report['4.0']['precision']*100,2))+\"           \"+\n",
    "                                      str(np.round(report['4.0']['recall']*100,2)) +\"       \"+\n",
    "                                      str(np.round(report['4.0']['f1-score']*100,2)))\n",
    "print(\"Severe dementia             \"+ str(np.round(report['5.0']['precision']*100,2))+\"           \"+\n",
    "                                      str(np.round(report['5.0']['recall']*100,2)) +\"       \"+\n",
    "                                      str(np.round(report['5.0']['f1-score']*100,2)))\n",
    "print()\n",
    "print(\"Accuracy on test data: %0.2f\" %np.round(np.mean(test_score)*100,2)+'%')\n",
    "print(\"Accuracy on training data: %0.2f\" %np.round(np.mean(train_score)*100,2)+'%')\n",
    "print()\n",
    "print(\"MSE Loss: %0.3f\" %np.mean(loss))\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.5 Plot prediction results and Model performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot confusion matrix\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "fig = plt.figure(figsize = (7, 7))\n",
    "plt.rc('font', size=15) \n",
    "ax = plt.axes()\n",
    "ax = sns.heatmap(CM, annot=True,  fmt='', cmap='Blues')\n",
    "\n",
    "ax.set_title('Confusion Matrix: \\nClinical Dementia Rating (CDR) Predictions (%)\\n');\n",
    "ax.set_xlabel('Predicted CDR')\n",
    "ax.set_ylabel('Actual CDR');\n",
    "\n",
    "## Ticket labels - List must be in alphabetical order\n",
    "ax.xaxis.set_ticklabels(['(0) No dementia' , '(0.5) Questionable\\nimpairment', '(1) Mild\\ndementia', '(2) Moderate\\ndementia' , '(3) Severe\\ndementia'])\n",
    "ax.yaxis.set_ticklabels(['(0) No dementia' , '(0.5) Questionable\\nimpairment', '(1) Mild\\ndementia', '(2) Moderate\\ndementia' , '(3) Severe\\ndementia'])\n",
    "\n",
    "plt.setp(ax.get_yticklabels(), rotation=0, ha=\"right\",rotation_mode=\"anchor\")\n",
    "plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",rotation_mode=\"anchor\")\n",
    "\n",
    "\n",
    "## Display the visualization of the Confusion Matrix.\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Model Performance\n",
    "precision_bar = [np.round(report['1.0']['precision']*100,2), np.round(report['2.0']['precision']*100,2),\n",
    "                 np.round(report['3.0']['precision']*100,2), np.round(report['4.0']['precision']*100,2), \n",
    "                 np.round(report['5.0']['precision']*100,2)]\n",
    "\n",
    "recall_bar =  [np.round(report['1.0']['recall']*100,2), np.round(report['2.0']['recall']*100,2),\n",
    "               np.round(report['3.0']['recall']*100,2), np.round(report['4.0']['recall']*100,2), \n",
    "               np.round(report['5.0']['recall']*100,2)]\n",
    "\n",
    "f1_bar =  [np.round(report['1.0']['f1-score']*100,2), np.round(report['2.0']['f1-score']*100,2),\n",
    "               np.round(report['3.0']['f1-score']*100,2), np.round(report['4.0']['f1-score']*100,2), \n",
    "               np.round(report['5.0']['f1-score']*100,2)]\n",
    "\n",
    "labels = ['No dementia\\n(0)' , 'Questionable\\nimpairment\\n(0.5) ', 'Mild\\ndementia\\n(1)', 'Moderate\\ndementia\\n(2)' , 'Severe\\ndementia\\n(3)']\n",
    "\n",
    "width =0.25\n",
    "\n",
    "\n",
    "fig = plt.figure(figsize = (14, 6))\n",
    "plt.rc('font', size=17) \n",
    "ax = plt.axes()\n",
    "\n",
    "plt.grid(color='#95a5a6', linestyle='--', linewidth=1, axis='y', alpha=0.7)\n",
    "b1 = plt.bar(np.arange(len(precision_bar))-width-0.03, precision_bar, width=width,color='cornflowerblue', label='Precision')\n",
    "b2 = plt.bar(np.arange(len(recall_bar)), recall_bar, width=width, color='peachpuff', label='Recall')\n",
    "b3 = plt.bar(np.arange(len(f1_bar))+ width+0.03, f1_bar, width=width, color='yellowgreen', label='f1-score')\n",
    "\n",
    "for bar in b1:\n",
    "    yval = bar.get_height()\n",
    "    plt.text(bar.get_x(), yval + 2, yval, fontsize=13)\n",
    "    \n",
    "for bar in b2:\n",
    "    yval = bar.get_height()\n",
    "    plt.text(bar.get_x(), yval + 2, yval, fontsize=13)\n",
    "    \n",
    "for bar in b3:\n",
    "    yval = bar.get_height()\n",
    "    plt.text(bar.get_x(), yval + 2, yval, fontsize=13)\n",
    "    \n",
    "plt.xticks(range(len(f1_bar)), labels)\n",
    "#plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",rotation_mode=\"anchor\")\n",
    "\n",
    "plt.xlabel('\\nCDR Class')\n",
    "plt.ylabel('Performance (%)')\n",
    "plt.title('Multiclass CDR Prediction Model Performance\\n')\n",
    "\n",
    "ax.legend(bbox_to_anchor=(1, 1),loc='upper left', title='Clinical Dementia Rating')\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}