Classify_complaint.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import cv2\n",
    "import os\n",
    "import tqdm\n",
    "from scipy.io import loadmat\n",
    "import matplotlib.pyplot as plt\n",
    "import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = []\n",
    "y_train = []\n",
    "data_path = \"C:/Users/Smit/janaagraha_dataset/images/\"\n",
    "classes = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [],
   "source": [
    "category = pd.read_csv('cd_bangalore_complaints.csv')\n",
    "imageid = pd.read_csv('cd_img_id_map.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "ls = pd.merge(category,imageid,on='generic_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ls = ls.drop(columns=['generic_id'])\n",
    "ls.drop(ls.columns[[0,1,2,3,4,5,7,8,9,10,11]], axis=1, inplace=True)\n",
    "ls['image_path'] = ls['image_path'].str.replace('images/','')\n",
    "ls['image_path'] = ls['image_path'].str.replace('images/','')\n",
    "ls = ls.dropna()\n",
    "ls['category_id'] = ls['category_id'].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "i=0\n",
    "for image in os.listdir(data_path):\n",
    "#     print(ls[ls['image_path']==image]['category_id'])\n",
    "    img = cv2.imread(data_path+\"/\"+image)\n",
    "    img = cv2.resize(img,(480,640))\n",
    "    X_train.append(img)\n",
    "    classes.append(ls[ls['image_path']==image]['category_id'])\n",
    "    i=i+1\n",
    "    if i==500:\n",
    "        break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "yes\n"
     ]
    }
   ],
   "source": [
    "xTrain = X_train\n",
    "X_train = []\n",
    "y_train = []\n",
    "N = len(classes)\n",
    "for i in  range(490):\n",
    "    if classes[i].empty:\n",
    "        print(\"yes\")\n",
    "        continue\n",
    "    X_train.append(xTrain[i])\n",
    "    y_train.append(classes[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(y_train)):\n",
    "    y_train[i]=int(y_train[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('cd_categories.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>parent_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Certificates</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>Crime and Safety</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>Electricity and Power Supply</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>Fire Safety</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>Garbage and Unsanitary Practices</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>188</th>\n",
       "      <td>421</td>\n",
       "      <td>Construction Of Sewage Treatment Plant</td>\n",
       "      <td>472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>189</th>\n",
       "      <td>87</td>\n",
       "      <td>Prevent Unauthorized Tree Cutting</td>\n",
       "      <td>473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>190</th>\n",
       "      <td>180</td>\n",
       "      <td>Defacement Of Trees</td>\n",
       "      <td>473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>191</th>\n",
       "      <td>355</td>\n",
       "      <td>Maintenance Of Trees/Saplings</td>\n",
       "      <td>473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>356</td>\n",
       "      <td>Plant Saplings/Trees in your locality</td>\n",
       "      <td>473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>193 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id                                   title  parent_id\n",
       "0      1                            Certificates          0\n",
       "1      2                        Crime and Safety          0\n",
       "2      4            Electricity and Power Supply          0\n",
       "3      5                             Fire Safety          0\n",
       "4      6        Garbage and Unsanitary Practices          0\n",
       "..   ...                                     ...        ...\n",
       "188  421  Construction Of Sewage Treatment Plant        472\n",
       "189   87       Prevent Unauthorized Tree Cutting        473\n",
       "190  180                     Defacement Of Trees        473\n",
       "191  355           Maintenance Of Trees/Saplings        473\n",
       "192  356   Plant Saplings/Trees in your locality        473\n",
       "\n",
       "[193 rows x 3 columns]"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "y=y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train=y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "ytrain = []\n",
    "classes = []\n",
    "uniq = np.unique(y_train)\n",
    "classes=[]\n",
    "for x in uniq:\n",
    "    classes.append(df[df['id']==int(x)]['title'].values)\n",
    "for x in y_train:\n",
    "    ytrain.append(df[df['id']==int(x)]['title'].values)\n",
    "y_train=[]\n",
    "for x in ytrain:\n",
    "    y_train.append(classes.index(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "489"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "X_train = np.asarray(X_train, dtype=np.float32)\n",
    "y_train = np.asarray(y_train, dtype=np.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(489, 640, 480, 3)\n",
      "(489, 1)\n"
     ]
    }
   ],
   "source": [
    "y_train = np.expand_dims(y_train, axis=1)\n",
    "print(X_train.shape)\n",
    "print(y_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train/255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train[:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.utils import np_utils\n",
    "#one hot encoder\n",
    "num_classes = len(np.unique(y_train))\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_4\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_16 (Conv2D)           (None, 640, 480, 16)      208       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_16 (MaxPooling (None, 320, 240, 16)      0         \n",
      "_________________________________________________________________\n",
      "conv2d_17 (Conv2D)           (None, 320, 240, 32)      2080      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_17 (MaxPooling (None, 160, 120, 32)      0         \n",
      "_________________________________________________________________\n",
      "conv2d_18 (Conv2D)           (None, 160, 120, 64)      8256      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_18 (MaxPooling (None, 80, 60, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_19 (Conv2D)           (None, 80, 60, 128)       32896     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_19 (MaxPooling (None, 40, 30, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_20 (Conv2D)           (None, 40, 30, 256)       131328    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_20 (MaxPooling (None, 20, 15, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 20, 15, 256)       0         \n",
      "_________________________________________________________________\n",
      "flatten_4 (Flatten)          (None, 76800)             0         \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 512)               39322112  \n",
      "_________________________________________________________________\n",
      "dropout_8 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 15)                7695      \n",
      "=================================================================\n",
      "Total params: 39,504,575\n",
      "Trainable params: 39,504,575\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Conv2D(filters=16,kernel_size=2,padding='same',activation='relu',\n",
    "                input_shape=(640,480,3)))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Conv2D(filters=32,kernel_size=2,padding='same',activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Conv2D(filters=64,kernel_size=2,padding='same',activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Conv2D(filters=128,kernel_size=2,padding='same',activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Conv2D(filters=256,kernel_size=2,padding='same',activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(512,activation='relu'))\n",
    "model.add(Dropout(0.4))\n",
    "model.add(Dense(15,activation='softmax'))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy', optimizer='rmsprop', \n",
    "                  metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      " - 142s - loss: 4.5373 - accuracy: 0.2577\n",
      "Epoch 2/5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Smit\\Anaconda2\\envs\\py36-test\\lib\\site-packages\\keras\\callbacks\\callbacks.py:707: RuntimeWarning: Can save best model only with val_loss available, skipping.\n",
      "  'skipping.' % (self.monitor), RuntimeWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 158s - loss: 1.9469 - accuracy: 0.4029\n",
      "Epoch 3/5\n",
      " - 155s - loss: 1.9685 - accuracy: 0.3722\n",
      "Epoch 4/5\n",
      " - 164s - loss: 1.8419 - accuracy: 0.4213\n",
      "Epoch 5/5\n",
      " - 181s - loss: 1.8374 - accuracy: 0.3885\n"
     ]
    }
   ],
   "source": [
    "from keras.callbacks import ModelCheckpoint   \n",
    "\n",
    "# train the model\n",
    "# checkpointer = ModelCheckpoint(filepath='mnist.model.best.hdf5',verbose=1,save_best_only=True)\n",
    "\n",
    "checkpointer = ModelCheckpoint(filepath='model.weights.best.hdf5', verbose=1, \n",
    "                               save_best_only=True)\n",
    "hist = model.fit(X_train, y_train, batch_size=32, epochs=5, callbacks=[checkpointer], \n",
    "          verbose=2, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "test=[]\n",
    "img = cv2.imread(\"C:\\\\Users\\\\Smit\\\\janaagraha_dataset\\\\images\\\\original_100203.jpg\")\n",
    "plt.imshow(img)\n",
    "plt.show()\n",
    "img = cv2.resize(img,(480,640))\n",
    "test.append(img)\n",
    "test = np.asarray(test, dtype=np.float32)\n",
    "test = test/255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Garbage and Unsanitary Practices']\n"
     ]
    }
   ],
   "source": [
    "classimg = model.predict_classes(test)\n",
    "print(classes[classimg[0]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1], dtype=int64)"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}