models/FFNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "import numpy as np\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation\n",
    "from datetime import datetime\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import train_test_split\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from utils import *\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_day_load_data(load_path):\n",
    "    '''read csv data at load_path and return it in a numpy matrix'''\n",
    "    data = pd.read_csv(load_path)\n",
    "    date_ = load_path.split('/')[-1][:-4] # get date of '01-10-2017' format\n",
    "    data['date'] = datetime.strptime(date_, \"%d-%m-%Y\").date()\n",
    "    data['time'] += ':00'\n",
    "    data['datetime'] = pd.to_timedelta(data['time'])\n",
    "    data['hour'] = data['datetime'].dt.seconds // 3600 \n",
    "    return data.as_matrix()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_30min_wise_data(day_load_data, day_whether_data):\n",
    "    '''Returns time, day temp, humidity and load values for each 30mins in a numpy matrix'''\n",
    "    data = []\n",
    "    for i in range(0, 288, 6): # SLDC load data has time step of 5 mins, so total 24 * (60 // 5)\n",
    "        load = day_load_data[i][1]\n",
    "        day = day_load_data[i][2].weekday()\n",
    "        \n",
    "        try:\n",
    "            humidity = float(day_whether_data[i//6][1].replace('%', ''))\n",
    "        except Exception as e:\n",
    "            humidity = 50 # some dirt values in data\n",
    "        try:\n",
    "            temp = float(day_whether_data[i//6][0]) # i//6 because whether data is already in 30 min step size\n",
    "        except Exception as err:\n",
    "            temp = 28 # aiyvi\n",
    "        data.append([i, day, temp, humidity, load])\n",
    "    return np.asarray(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_data(data_list, shuffle=True):\n",
    "    '''Return data of all csv files in SLDC_Data/ folder in matrix form'''\n",
    "    data = []\n",
    "    for load_path in data_list:\n",
    "        try:\n",
    "            whether_path = load_path.replace('SLDC_Data', 'Whether_Data')\n",
    "            day_load_data = get_day_load_data(load_path)\n",
    "            day_whether_data = pd.read_csv(whether_path)\n",
    "            day_data = get_30min_wise_data(day_load_data, day_whether_data[['Temp.', 'Humidity']].as_matrix())\n",
    "            data.append(day_data)\n",
    "        except Exception as e:\n",
    "            print(load_path, e)\n",
    "    data = np.asarray(data) # (115, 48, 5)\n",
    "    data = data.reshape(-1, 5) # (5520, 5)\n",
    "    if shuffle: np.random.shuffle(data)\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SLDC_Data/2017/12/07-12-2017.csv index 132 is out of bounds for axis 0 with size 132\n",
      "SLDC_Data/2018/01/04-01-2018.csv index 276 is out of bounds for axis 0 with size 276\n"
     ]
    }
   ],
   "source": [
    "# data preprocessing\n",
    "data_list = glob.glob('SLDC_Data/*/*/*.csv')\n",
    "train, test = train_test_split(data_list, test_size=0.1)\n",
    "train_data = get_data(train)\n",
    "X_train = train_data[:, :-1] # features\n",
    "Y_train = train_data[:, -1].reshape(-1, 1) # load, reshaped to 2D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# rescale the data to 0-1 values`\n",
    "feature_scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "load_scaler = MinMaxScaler(feature_range=(0, 1)) # try StandardScaler()\n",
    "X_train = feature_scaler.fit_transform(X_train)\n",
    "Y_train = load_scaler.fit_transform(Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((6048, 4), (6048, 1), (48, 4), (48, 1))"
      ]
     },
     "execution_count": 301,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, Y_train.shape, X_test.shape, Y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def getModel():\n",
    "    model = Sequential()\n",
    "    model.add(Dense(256, input_shape=(4, )))\n",
    "    model.add(Activation('relu'))\n",
    "    model.add(Dense(256))\n",
    "    model.add(Activation('relu'))\n",
    "    model.add(Dense(256))\n",
    "    model.add(Activation('relu'))\n",
    "    model.add(Dense(1, activation='sigmoid'))\n",
    "    model.compile(loss='mse', optimizer='adam')\n",
    "    print(model.summary())\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_40 (Dense)             (None, 256)               1280      \n",
      "_________________________________________________________________\n",
      "activation_30 (Activation)   (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_41 (Dense)             (None, 256)               65792     \n",
      "_________________________________________________________________\n",
      "activation_31 (Activation)   (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_42 (Dense)             (None, 256)               65792     \n",
      "_________________________________________________________________\n",
      "activation_32 (Activation)   (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_43 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 133,121\n",
      "Trainable params: 133,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "model = getModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 5443 samples, validate on 605 samples\n",
      "Epoch 1/50\n",
      "5443/5443 [==============================] - 2s 285us/step - loss: 0.0202 - val_loss: 0.0132\n",
      "Epoch 2/50\n",
      "5443/5443 [==============================] - 1s 174us/step - loss: 0.0087 - val_loss: 0.0079\n",
      "Epoch 3/50\n",
      "5443/5443 [==============================] - 1s 191us/step - loss: 0.0067 - val_loss: 0.0072\n",
      "Epoch 4/50\n",
      "5443/5443 [==============================] - 1s 180us/step - loss: 0.0062 - val_loss: 0.0066\n",
      "Epoch 5/50\n",
      "5443/5443 [==============================] - 1s 214us/step - loss: 0.0057 - val_loss: 0.0070\n",
      "Epoch 6/50\n",
      "5443/5443 [==============================] - 1s 188us/step - loss: 0.0058 - val_loss: 0.0062\n",
      "Epoch 7/50\n",
      "5443/5443 [==============================] - 1s 190us/step - loss: 0.0053 - val_loss: 0.0067\n",
      "Epoch 8/50\n",
      "5443/5443 [==============================] - 1s 189us/step - loss: 0.0052 - val_loss: 0.0070\n",
      "Epoch 9/50\n",
      "5443/5443 [==============================] - 1s 187us/step - loss: 0.0049 - val_loss: 0.0072\n",
      "Epoch 10/50\n",
      "5443/5443 [==============================] - 1s 195us/step - loss: 0.0051 - val_loss: 0.0063\n",
      "Epoch 11/50\n",
      "5443/5443 [==============================] - 1s 187us/step - loss: 0.0047 - val_loss: 0.0057\n",
      "Epoch 12/50\n",
      "5443/5443 [==============================] - 1s 193us/step - loss: 0.0049 - val_loss: 0.0057\n",
      "Epoch 13/50\n",
      "5443/5443 [==============================] - 1s 174us/step - loss: 0.0046 - val_loss: 0.0061\n",
      "Epoch 14/50\n",
      "5443/5443 [==============================] - 1s 203us/step - loss: 0.0049 - val_loss: 0.0055\n",
      "Epoch 15/50\n",
      "5443/5443 [==============================] - 1s 193us/step - loss: 0.0044 - val_loss: 0.0061\n",
      "Epoch 16/50\n",
      "5443/5443 [==============================] - 1s 204us/step - loss: 0.0043 - val_loss: 0.0053\n",
      "Epoch 17/50\n",
      "5443/5443 [==============================] - 1s 178us/step - loss: 0.0043 - val_loss: 0.0051\n",
      "Epoch 18/50\n",
      "5443/5443 [==============================] - 1s 202us/step - loss: 0.0042 - val_loss: 0.0061\n",
      "Epoch 19/50\n",
      "5443/5443 [==============================] - 1s 182us/step - loss: 0.0040 - val_loss: 0.0051\n",
      "Epoch 20/50\n",
      "5443/5443 [==============================] - 1s 196us/step - loss: 0.0041 - val_loss: 0.0048\n",
      "Epoch 21/50\n",
      "5443/5443 [==============================] - 1s 180us/step - loss: 0.0044 - val_loss: 0.0048\n",
      "Epoch 22/50\n",
      "5443/5443 [==============================] - 1s 195us/step - loss: 0.0041 - val_loss: 0.0044\n",
      "Epoch 23/50\n",
      "5443/5443 [==============================] - 1s 180us/step - loss: 0.0041 - val_loss: 0.0046\n",
      "Epoch 24/50\n",
      "5443/5443 [==============================] - 1s 196us/step - loss: 0.0039 - val_loss: 0.0048\n",
      "Epoch 25/50\n",
      "5443/5443 [==============================] - 1s 198us/step - loss: 0.0039 - val_loss: 0.0046\n",
      "Epoch 26/50\n",
      "5443/5443 [==============================] - 1s 201us/step - loss: 0.0041 - val_loss: 0.0052\n",
      "Epoch 27/50\n",
      "5443/5443 [==============================] - 1s 177us/step - loss: 0.0041 - val_loss: 0.0046\n",
      "Epoch 28/50\n",
      "5443/5443 [==============================] - 1s 205us/step - loss: 0.0038 - val_loss: 0.0053\n",
      "Epoch 29/50\n",
      "5443/5443 [==============================] - 1s 180us/step - loss: 0.0038 - val_loss: 0.0045\n",
      "Epoch 30/50\n",
      "5443/5443 [==============================] - 1s 198us/step - loss: 0.0037 - val_loss: 0.0049\n",
      "Epoch 31/50\n",
      "5443/5443 [==============================] - 1s 189us/step - loss: 0.0038 - val_loss: 0.0047\n",
      "Epoch 32/50\n",
      "5443/5443 [==============================] - 1s 193us/step - loss: 0.0037 - val_loss: 0.0050\n",
      "Epoch 33/50\n",
      "5443/5443 [==============================] - 1s 194us/step - loss: 0.0038 - val_loss: 0.0047\n",
      "Epoch 34/50\n",
      "5443/5443 [==============================] - 1s 210us/step - loss: 0.0037 - val_loss: 0.0045\n",
      "Epoch 35/50\n",
      "5443/5443 [==============================] - 1s 249us/step - loss: 0.0037 - val_loss: 0.0051\n",
      "Epoch 36/50\n",
      "5443/5443 [==============================] - 1s 214us/step - loss: 0.0036 - val_loss: 0.0052\n",
      "Epoch 37/50\n",
      "5443/5443 [==============================] - 1s 210us/step - loss: 0.0038 - val_loss: 0.0046\n",
      "Epoch 38/50\n",
      "5443/5443 [==============================] - 1s 189us/step - loss: 0.0036 - val_loss: 0.0043\n",
      "Epoch 39/50\n",
      "5443/5443 [==============================] - 1s 197us/step - loss: 0.0035 - val_loss: 0.0050\n",
      "Epoch 40/50\n",
      "5443/5443 [==============================] - 1s 196us/step - loss: 0.0037 - val_loss: 0.0042\n",
      "Epoch 41/50\n",
      "5443/5443 [==============================] - 1s 211us/step - loss: 0.0036 - val_loss: 0.0043\n",
      "Epoch 42/50\n",
      "5443/5443 [==============================] - 1s 230us/step - loss: 0.0034 - val_loss: 0.0044\n",
      "Epoch 43/50\n",
      "5443/5443 [==============================] - 1s 210us/step - loss: 0.0035 - val_loss: 0.0040\n",
      "Epoch 44/50\n",
      "5443/5443 [==============================] - 1s 229us/step - loss: 0.0035 - val_loss: 0.0045\n",
      "Epoch 45/50\n",
      "5443/5443 [==============================] - 1s 188us/step - loss: 0.0035 - val_loss: 0.0047\n",
      "Epoch 46/50\n",
      "5443/5443 [==============================] - 1s 204us/step - loss: 0.0035 - val_loss: 0.0041\n",
      "Epoch 47/50\n",
      "5443/5443 [==============================] - 1s 200us/step - loss: 0.0033 - val_loss: 0.0049\n",
      "Epoch 48/50\n",
      "5443/5443 [==============================] - 1s 214us/step - loss: 0.0033 - val_loss: 0.0051\n",
      "Epoch 49/50\n",
      "5443/5443 [==============================] - 1s 228us/step - loss: 0.0035 - val_loss: 0.0041\n",
      "Epoch 50/50\n",
      "5443/5443 [==============================] - 1s 215us/step - loss: 0.0034 - val_loss: 0.0050\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f0ad65cdbe0>"
      ]
     },
     "execution_count": 333,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, Y_train, batch_size=32, epochs=50, verbose=1, validation_split=0.1, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def predict(data):\n",
    "    '''selects a date in data and plots the predicted curve using the trained model'''\n",
    "    try:\n",
    "        x = np.random.randint(len(data)) # to select a date in data data\n",
    "        data_data = get_data([data[x]], shuffle=False)\n",
    "        X_data = data_data[:, :-1] # features\n",
    "        Y_data = data_data[:, -1].reshape(-1, 1) # load, reshaped to 2D\n",
    "\n",
    "        X_data = feature_scaler.fit_transform(X_data)\n",
    "        Y_data = load_scaler.fit_transform(Y_data)\n",
    "\n",
    "        preds = model.predict(X_data)\n",
    "        preds = load_scaler.inverse_transform(preds)\n",
    "        Y_ = load_scaler.inverse_transform(Y_data)\n",
    "\n",
    "        plt.plot(range(len(Y_)), Y_)\n",
    "        plt.plot(range(len(preds)), preds)\n",
    "        plt.legend(['actual', 'pred'])\n",
    "        plt.title(data[x])\n",
    "        plt.show()\n",
    "    except Exception as e:\n",
    "        print('Err in data', e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEICAYAAABWJCMKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlCUAY4zxKEsAxhjjUZYAjDHGo/4PYV4eOMXJVZsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0ad704c390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# prediction on test data\n",
    "predict(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 354,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEICAYAAAC55kg0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0ad745a668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# prediction on train data\n",
    "predict(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1957 2007\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0ad7e35c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# prediction on train data as a whole\n",
    "\n",
    "preds = model.predict(X_train)\n",
    "preds = load_scaler.inverse_transform(preds)\n",
    "Y_ = load_scaler.inverse_transform(Y_train)\n",
    "\n",
    "x = np.random.randint(len(preds))\n",
    "x_max = 50\n",
    "print(x,x+x_max)\n",
    "plt.plot(range(x, x+x_max), preds[x:x+x_max])\n",
    "plt.plot(range(x, x+x_max), Y_[x:x+x_max])\n",
    "plt.legend(['pred', 'actual'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ML",
   "language": "python",
   "name": "ml"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  },
  "widgets": {
   "state": {},
   "version": "1.1.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}