svm[pavia university].ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from sklearn import decomposition\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.decomposition import PCA\n",
    "import seaborn as sns\n",
    "from scipy.io import loadmat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = loadmat('PaviaU.mat')['paviaU']\n",
    "y = loadmat('PaviaU_gt.mat')['paviaU_gt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X: (610, 340, 103)\n",
      "y: (610, 340)\n"
     ]
    }
   ],
   "source": [
    "print(\"X:\", (X.shape))\n",
    "print(\"y:\", (y.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=uint8)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 10))\n",
    "plt.imshow(y, cmap='jet')\n",
    "plt.colorbar()\n",
    "plt.axis('off')\n",
    "plt.title('Ground Truth')\n",
    "plt.gca().legend(['0', '1', '2', '3', '4', '5', '6', '7', '8', '9'], loc='upper left')\n",
    "#plt.savefig('ground_truth.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = X.reshape(X.shape[0]*X.shape[1], X.shape[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = y.ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X: (207400, 103)\n",
      "y: (207400,)\n"
     ]
    }
   ],
   "source": [
    "print(\"X:\", (X.shape))\n",
    "print(\"y:\", (y.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2 3 4 5 6 7 8 9]\n",
      "[164624   6631  18649   2099   3064   1345   5029   1330   3682    947]\n"
     ]
    }
   ],
   "source": [
    "label,counts=np.unique(y,return_counts=True)\n",
    "print(label)\n",
    "print(counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Standardizing the features\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler_model = StandardScaler()\n",
    "scaler_model.fit(X.astype(float))\n",
    "X=scaler_model.transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(207400, 103)\n",
      "(207400,)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        0\n",
      "0  123468\n",
      "1    4973\n",
      "2   13987\n",
      "3    1574\n",
      "4    2298\n",
      "5    1009\n",
      "6    3772\n",
      "7     997\n",
      "8    2762\n",
      "9     710\n",
      "       0\n",
      "0  41156\n",
      "1   1658\n",
      "2   4662\n",
      "3    525\n",
      "4    766\n",
      "5    336\n",
      "6   1257\n",
      "7    333\n",
      "8    920\n",
      "9    237\n"
     ]
    }
   ],
   "source": [
    "label_tr,counts_tr=np.unique(y_train,return_counts=True)\n",
    "label_te,counts_te=np.unique(y_test,return_counts=True)\n",
    "print(pd.DataFrame(counts_tr,label_tr))\n",
    "print(pd.DataFrame(counts_te,label_te))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state = 10, test_size = 0.25, stratify=y )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(155550, 103)\n",
      "(51850, 103)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Applying Scalar to train and test Dataset\n",
    "\n",
    "from sklearn import preprocessing\n",
    "scaler = preprocessing.StandardScaler().fit(X_train) \n",
    "X_train= scaler.transform(X_train)\n",
    "X_test=scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Appplying SVM \n",
    "\n",
    "from sklearn.svm import SVC\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = SVC(C=100,gamma=0.01,decision_function_shape='ovo')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=100, cache_size=200, class_weight=None, coef0=0.0,\n",
       "    decision_function_shape='ovo', degree=3, gamma=0.01, kernel='rbf',\n",
       "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "    tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = clf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy : 84.17164898746384\n"
     ]
    }
   ],
   "source": [
    "print(\"Accuracy :\",metrics.accuracy_score(y_pred, y_test)*100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cohen Kappa Score : 43.861578083437614\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import cohen_kappa_score\n",
    "print(\"Cohen Kappa Score :\",cohen_kappa_score(y_pred, y_test)*100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\THIS PC\\Anaconda3\\lib\\site-packages\\sklearn\\metrics\\classification.py:1437: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.86      0.96      0.91     41156\n",
      "           1       0.20      0.00      0.00      1658\n",
      "           2       0.73      0.64      0.68      4662\n",
      "           3       0.65      0.30      0.41       525\n",
      "           4       0.70      0.11      0.19       766\n",
      "           5       0.82      0.91      0.86       336\n",
      "           6       0.89      0.01      0.01      1257\n",
      "           7       0.68      0.68      0.68       333\n",
      "           8       0.66      0.20      0.31       920\n",
      "           9       0.00      0.00      0.00       237\n",
      "\n",
      "    accuracy                           0.84     51850\n",
      "   macro avg       0.62      0.38      0.41     51850\n",
      "weighted avg       0.81      0.84      0.80     51850\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix,classification_report\n",
    "mat = confusion_matrix(y_pred, y_test)\n",
    "pd.DataFrame(mat)\n",
    "# print(confusion_matrix(y_test, y_pred ))\n",
    "print(classification_report(y_test, y_pred ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       0     1     2    3    4    5     6    7    8    9\n",
      "0  39695  1653  1677  365  681   30  1244  108  736  237\n",
      "1      4     1     0    0    0    0     0    0    0    0\n",
      "2   1080     0  2985    0    1    0     5    0    0    0\n",
      "3     82     0     0  155    0    0     0    0    0    0\n",
      "4     36     0     0    0   84    0     0    0    0    0\n",
      "5     66     0     0    0    0  306     0    0    0    0\n",
      "6      1     0     0    0    0    0     8    0    0    0\n",
      "7    101     4     0    0    0    0     0  225    0    0\n",
      "8     91     0     0    5    0    0     0    0  184    0\n",
      "9      0     0     0    0    0    0     0    0    0    0\n"
     ]
    }
   ],
   "source": [
    "print(pd.DataFrame(mat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}