TimeSeries_Forecasting.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2f4526c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from datetime import datetime as dt\n",
    "\n",
    "\n",
    "# Synthetic dataset\n",
    "from sklearn.datasets import make_classification# Data processing\n",
    "from statsmodels.graphics.tsaplots import plot_acf\n",
    "from statsmodels.graphics.tsaplots import plot_pacf\n",
    "from statsmodels.tsa.stattools import pacf\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from collections import Counter\n",
    "\n",
    "import plotly.graph_objects as go\n",
    "\n",
    "# Visualization\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# Model and performance\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers, losses\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report, mean_squared_error\n",
    "from keras.utils.vis_utils import plot_model\n",
    "\n",
    "import scipy.cluster.hierarchy as sch\n",
    "\n",
    "from sklearn.cluster import DBSCAN\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.metrics import pairwise_distances\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "import xgboost as xgb\n",
    "from sklearn.multioutput import MultiOutputRegressor\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.metrics import mean_absolute_percentage_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "06ce39ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset1 = pd.read_csv('C:/Users/aggelosuser/Desktop/test9after.csv', parse_dates=True)#, index_col=\"time\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "43971066",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset1 = dataset1.drop([\"PS1_V\"], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "bc71d1d4",
   "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>MEM_USAGE</th>\n",
       "      <th>CPU_USAGE</th>\n",
       "      <th>TEMP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26.505572</td>\n",
       "      <td>6.800000</td>\n",
       "      <td>28.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>26.505572</td>\n",
       "      <td>7.200000</td>\n",
       "      <td>28.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>26.505572</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>28.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>26.505572</td>\n",
       "      <td>14.960630</td>\n",
       "      <td>28.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26.505572</td>\n",
       "      <td>9.225092</td>\n",
       "      <td>28.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15458</th>\n",
       "      <td>25.602151</td>\n",
       "      <td>68.131868</td>\n",
       "      <td>32.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15459</th>\n",
       "      <td>25.602151</td>\n",
       "      <td>68.131868</td>\n",
       "      <td>32.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15460</th>\n",
       "      <td>25.602151</td>\n",
       "      <td>68.131868</td>\n",
       "      <td>32.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15461</th>\n",
       "      <td>25.602151</td>\n",
       "      <td>68.131868</td>\n",
       "      <td>32.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15462</th>\n",
       "      <td>25.602151</td>\n",
       "      <td>68.131868</td>\n",
       "      <td>32.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15463 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       MEM_USAGE  CPU_USAGE   TEMP\n",
       "0      26.505572   6.800000  28.25\n",
       "1      26.505572   7.200000  28.25\n",
       "2      26.505572  42.000000  28.25\n",
       "3      26.505572  14.960630  28.25\n",
       "4      26.505572   9.225092  28.25\n",
       "...          ...        ...    ...\n",
       "15458  25.602151  68.131868  32.00\n",
       "15459  25.602151  68.131868  32.00\n",
       "15460  25.602151  68.131868  32.00\n",
       "15461  25.602151  68.131868  32.00\n",
       "15462  25.602151  68.131868  32.00\n",
       "\n",
       "[15463 rows x 3 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b084c174",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = dataset1.iloc[:3150]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "37ec4528",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataset.plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d65a267c",
   "metadata": {},
   "source": [
    "XGBOOST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "id": "03ab2252",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 1.796911\n",
      "RMSE: 1.340489\n",
      "MAE: 0.448495\n",
      "MAPE: 0.015917%\n",
      "SMAPE:0.016448\n",
      "R-squared:0.703249\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.6332480907440186\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.time()\n",
    "#%%timeit\n",
    "#result = %time \n",
    "#import timeit\n",
    "#%timeit\n",
    "#a = (timeit.timeit())\n",
    "#print(a)\n",
    "\n",
    "from numpy import asarray\n",
    "from pandas import read_csv\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from xgboost import XGBRegressor\n",
    "from matplotlib import pyplot\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "# transform a time series dataset into a supervised learning dataset\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\tn_vars = 1 if type(data) is list else data.shape[0]\n",
    "\tdf = DataFrame(data)\n",
    "\tcols = list()\n",
    "\t# input sequence (t-n, ... t-1)\n",
    "\tfor i in range(n_in, 0, -1):\n",
    "\t\tcols.append(df.shift(i))\n",
    "\t# forecast sequence (t, t+1, ... t+n)\n",
    "\tfor i in range(0, n_out):\n",
    "\t\tcols.append(df.shift(-i))\n",
    "\t# put it together\n",
    "\tagg = concat(cols, axis=1)\n",
    "\t# drop rows with NaN values\n",
    "\tif dropnan:\n",
    "\t\tagg.dropna(inplace=True)\n",
    "\treturn agg.values\n",
    "\n",
    "# split dataset into train/test sets\n",
    "def train_test_split(data, n_test):\n",
    "\treturn data[:-n_test, :], data[-n_test:, :]\n",
    "\n",
    "# fit xgboost and make a one step prediction\n",
    "def xgboost_forecast(train, testX):\n",
    "\ttrain = asarray(train)\n",
    "\t# split into input and output columns\n",
    "\ttrainX, trainy = train[:, :-1], train[:, -1]\n",
    "\t# fit model\n",
    "\t#model = XGBRegressor(objective='reg:squarederror', n_estimators=1000)\n",
    "\tmodel = XGBRegressor(objective='reg:squarederror', n_estimators=100, max_depth=4, eta=0.1, \n",
    "\tmin_child_weight=1,subsample=0.9, colsample_bytree=0.9)\n",
    "\tmodel.fit(trainX, trainy)\n",
    "    #multi-step prediction\n",
    "\t#trained_model = MultiOutputRegressor(model).fit(trainX, trainy)\n",
    "    #train_forecasts = trained_model.predict(x_train)\n",
    "\t#print(testX)\n",
    "\tyhat = model.predict(asarray([testX]))\n",
    "\t#print(yhat)\n",
    "\treturn yhat[0]\n",
    "\n",
    "# walk-forward validation for univariate data\n",
    "def walk_forward_validation(data, n_test):\n",
    "\tpredictions = list()\n",
    "\t\n",
    "\ttrain, test = train_test_split(data, n_test)\n",
    "\t# seed history with training dataset\n",
    "\thistory = [x for x in train]\n",
    "\t# step over each time-step in the test set\n",
    "\tfor i in range(len(test)):\n",
    "\t\t# split test row into input and output columns\n",
    "\t\ttestX, testy = test[i, :-1], test[i, -1]\n",
    "\t\t\n",
    "\t\tyhat = xgboost_forecast(history, testX)\n",
    "\t\t# store forecast in list of predictions\n",
    "\t\tpredictions.append(yhat)\n",
    "\t\t# add actual observation to history for the next loop\n",
    "\t\thistory.append(test[i])\n",
    "\t\t#print('>expected=%.1f, predicted=%.1f' % (testy, yhat))\n",
    "\t# prediction error\n",
    "\terror = mean_absolute_error(test[:, -1], predictions)\n",
    "\treturn error, test[:, -1], predictions\n",
    "\n",
    "\n",
    "# transform the time series data into supervised learning\n",
    "data = series_to_supervised(dataset['TEMP'], n_in=8)\n",
    "#print(data.shape)\n",
    "yhat = np.array(yhat)\n",
    "# evaluate\n",
    "mae, y, yhat = walk_forward_validation(data, 100)\n",
    "#print('MAE: %.3f' % mae)\n",
    "# plot expected vs preducted\n",
    "pyplot.plot(y, label='Expected')\n",
    "pyplot.plot(yhat, label='Predicted')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n",
    "\n",
    "mse = mean_squared_error(y, yhat)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(y, yhat)\n",
    "mape = np.mean(np.abs((y - yhat) / y) * 1)\n",
    "print('MSE: {:.6f}'.format(mse))\n",
    "print('RMSE: {:.6f}'.format(rmse))\n",
    "print('MAE: {:.6f}'.format(mae))\n",
    "print('MAPE: {:.6f}%'.format(mape))\n",
    "def smape(actual, forecast):\n",
    "    \"\"\"Symmetric mean absolute percentage error\"\"\"\n",
    "    return np.mean((np.abs(actual - forecast) * 2 / (np.abs(actual) + np.abs(forecast))))\n",
    "smape_score = smape(y, yhat)\n",
    "print(\"SMAPE:{:.6f}\" .format(smape_score))\n",
    "# Calculate R-squared score\n",
    "r2 = r2_score(y,yhat)\n",
    "print(\"R-squared:{:.6f}\" .format(r2))\n",
    "#plot forecasts against actual outcomes\n",
    "pyplot.plot(y)\n",
    "pyplot.plot(yhat, color='red')\n",
    "pyplot.show()\n",
    "#print(result)\n",
    "#%time \n",
    "#print(timeit.timeit() - a)\n",
    "end = time.time()\n",
    "print(end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "id": "8f2cfd5c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3145, 6)\n",
      "MAE: 0.742\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 2.992858\n",
      "RMSE: 1.729988\n",
      "MAE: 0.742301\n",
      "MAPE: 0.027374%\n",
      "SMAPE:0.028026\n",
      "R-squared:0.505744\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.150763034820557\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.time()\n",
    "\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\t\"\"\"\n",
    "\tFrame a time series as a supervised learning dataset.\n",
    "\tArguments:\n",
    "\t\tdata: Sequence of observations as a list or NumPy array.\n",
    "\t\tn_in: Number of lag observations as input (X).\n",
    "\t\tn_out: Number of observations as output (y).\n",
    "\t\tdropnan: Boolean whether or not to drop rows with NaN values.\n",
    "\tReturns:\n",
    "\t\tPandas DataFrame of series framed for supervised learning.\n",
    "\t\"\"\"\n",
    "# \tn_vars = 1 if type(data) is list else data.shape[0]\n",
    "# \tdf = DataFrame(data)\n",
    "# \tcols = list(), list()\n",
    "# \t# input sequence (t-n, ... t-1)\n",
    "# \tfor i in range(n_in, 0, -1):\n",
    "# \t\tcols.append(df.shift(i))\n",
    "# \t\t# forecast sequence (t, t+1, ... t+n)\n",
    "# \tfor i in range(0, n_out):\n",
    "# \t\tcols.append(df.shift(-i))\n",
    "# \t# put it all together\n",
    "# \tagg = concat(cols, axis=1)\n",
    "# \t# drop rows with NaN values\n",
    "# \tif dropnan:\n",
    "# \t\tagg.dropna(inplace=True)\n",
    "# \treturn agg.values\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\tn_vars = 1 if type(data) is list else data.shape[0]\n",
    "\tdf = DataFrame(data)\n",
    "\tcols = list()\n",
    "\t# input sequence (t-n, ... t-1)\n",
    "\tfor i in range(n_in, 0, -1):\n",
    "\t\tcols.append(df.shift(i))\n",
    "\t# forecast sequence (t, t+1, ... t+n)\n",
    "\tfor i in range(0, n_out):\n",
    "\t\tcols.append(df.shift(-i))\n",
    "\t# put it all together\n",
    "\tagg = concat(cols, axis=1)\n",
    "\t# drop rows with NaN values\n",
    "\tif dropnan:\n",
    "\t\tagg.dropna(inplace=True)\n",
    "\treturn agg.values\n",
    "\n",
    "\n",
    "# split a univariate dataset into train/test sets\n",
    "def train_test_split(data, n_test):\n",
    "\treturn data[:-n_test, :], data[-n_test:, :]\n",
    "\n",
    "# fit an xgboost model and make a one step prediction\n",
    "def xgboost_forecast(train, testX):\n",
    "\t# transform list into array\n",
    "\ttrain = asarray(train)\n",
    "\t# split into input and output columns\n",
    "\ttrainX, trainy = train[:, :-2], train[:, -2:]\n",
    "\t# fit model\n",
    "\t#model = XGBRegressor(objective='reg:squarederror', n_estimators=1000)\n",
    "\t#print(train.shape)\n",
    "\t#print(trainX.shape)\n",
    "\t#print(trainy.shape)\n",
    "    # fitting\n",
    "\t#model = MultiOutputRegressor(xgb.XGBRegressor(objective='reg:squarederror', n_estimators=1000)).fit(trainX, trainy)\n",
    "\tmodel = MultiOutputRegressor(xgb.XGBRegressor(objective='reg:squarederror', n_estimators=100, max_depth=4, eta=0.1, \n",
    "\tmin_child_weight=1,subsample=0.9, colsample_bytree=0.9)).fit(trainX, trainy)\n",
    "    # predicting\n",
    "\n",
    "    #print(np.mean((multioutputregressor.predict(X) - y)**2, axis=0))\n",
    "\t#model.fit(trainX, trainy)\n",
    "\t# make a one-step prediction\n",
    "\t#print(testX)\n",
    "\t#model = MultiOutputRegressor(model).fit(trainX, trainy)\n",
    "\tyhat = model.predict(asarray([testX]))\n",
    "\t#print(\"yhat is :\",yhat)\n",
    "\treturn yhat[0]\n",
    "\n",
    "# walk-forward validation for univariate data\n",
    "def walk_forward_validation(data, n_test):\n",
    "\tpredictions = list()\n",
    "\t# split dataset\n",
    "\ttrain, test = train_test_split(data, n_test)\n",
    "\t# seed history with training dataset\n",
    "\thistory = [x for x in train]\n",
    "\t#print(\"history is \", history)\n",
    "\t# step over each time-step in the test set\n",
    "\tfor i in range(0,len(test),2):\n",
    "\t\t# split test row into input and output columns\n",
    "\t\ttestX, testy = test[i, :-2], test[i, -2:]\n",
    "\t\t#print(\"testX shape is:\", testX.shape)\n",
    "\t\t#print(\"testy shape is :\", testy.shape)\n",
    "\t\t# fit model on history and make a prediction\n",
    "\t\tyhat = xgboost_forecast(history, testX)\n",
    "\t\t# store forecast in list of predictions\n",
    "\t\tpredictions.append(yhat)\n",
    "\t\t# add actual observation to history for the next loop\n",
    "\t\thistory.append(test[i])\n",
    "\t\t# summarize progress\n",
    "\t\t#print('>expected=%.1f, predicted=%.1f' % (testy, yhat))\n",
    "\t\t#print('>expected = ',testy)\n",
    "\t\t#print('>predicted = ',yhat)       \n",
    "\t# estimate prediction error\n",
    "\t#print('>expected = ',testy)\n",
    "\tpredictions2 = [num for sublist in predictions for num in sublist]\n",
    "\t#print(len(predictions2))\n",
    "\t#print(test[:, -2:])\n",
    "\terror = mean_absolute_error(test[:, -1], predictions2)\n",
    "\treturn error, test[:, -1], predictions2\n",
    "\n",
    "\n",
    "# transform the time series data into supervised learning\n",
    "data = series_to_supervised(dataset['TEMP'],4,2)\n",
    "print(data.shape)\n",
    "# evaluate\n",
    "mae, y, yhat = walk_forward_validation(data, 100)\n",
    "#print(y)\n",
    "#print(yhat)\n",
    "print('MAE: %.3f' % mae)\n",
    "# plot expected vs preducted\n",
    "pyplot.plot(y, label='Expected')\n",
    "pyplot.plot(yhat, label='Predicted')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n",
    "\n",
    "mse = mean_squared_error(y, yhat)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(y, yhat)\n",
    "mape = np.mean(np.abs((y - yhat) / y) * 1)\n",
    "print('MSE: {:.6f}'.format(mse))\n",
    "print('RMSE: {:.6f}'.format(rmse))\n",
    "print('MAE: {:.6f}'.format(mae))\n",
    "print('MAPE: {:.6f}%'.format(mape))\n",
    "def smape(actual, forecast):\n",
    "    \"\"\"Symmetric mean absolute percentage error\"\"\"\n",
    "    return np.mean((np.abs(actual - forecast) * 2 / (np.abs(actual) + np.abs(forecast))))\n",
    "smape_score = smape(y, yhat)\n",
    "print(\"SMAPE:{:.6f}\" .format(smape_score))\n",
    "# Calculate R-squared score\n",
    "r2 = r2_score(y,yhat)\n",
    "print(\"R-squared:{:.6f}\" .format(r2))\n",
    "#plot forecasts against actual outcomes\n",
    "pyplot.plot(y)\n",
    "pyplot.plot(yhat, color='red')\n",
    "pyplot.show()\n",
    "\n",
    "end = time.time()\n",
    "print(end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "bd04a2db",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pyplot.plot(y, label='Expected')\n",
    "pyplot.plot(yhat, label='Predicted')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "65fbb4ea",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "b170a96b",
   "metadata": {},
   "source": [
    "MULTI_STEP FORECASTING"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "id": "dbde5198",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3071, 80)\n",
      "MAE: 3.207\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 13.406029\n",
      "RMSE: 3.661425\n",
      "MAE: 3.206860\n",
      "MAPE: 0.123073%\n",
      "SMAPE:0.132476\n",
      "R-squared:-1.213940\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8.982070922851562\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.time()\n",
    "\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\t\"\"\"\n",
    "\tFrame a time series as a supervised learning dataset.\n",
    "\tArguments:\n",
    "\t\tdata: Sequence of observations as a list or NumPy array.\n",
    "\t\tn_in: Number of lag observations as input (X).\n",
    "\t\tn_out: Number of observations as output (y).\n",
    "\t\tdropnan: Boolean whether or not to drop rows with NaN values.\n",
    "\tReturns:\n",
    "\t\tPandas DataFrame of series framed for supervised learning.\n",
    "\t\"\"\"\n",
    "\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\tn_vars = 1 if type(data) is list else data.shape[0]\n",
    "\tdf = DataFrame(data)\n",
    "\tcols = list()\n",
    "\t# input sequence (t-n, ... t-1)\n",
    "\tfor i in range(n_in, 0, -1):\n",
    "\t\tcols.append(df.shift(i))\n",
    "\t# forecast sequence (t, t+1, ... t+n)\n",
    "\tfor i in range(0, n_out):\n",
    "\t\tcols.append(df.shift(-i))\n",
    "\t# put it all together\n",
    "\tagg = concat(cols, axis=1)\n",
    "\t# drop rows with NaN values\n",
    "\tif dropnan:\n",
    "\t\tagg.dropna(inplace=True)\n",
    "\treturn agg.values\n",
    "\n",
    "\n",
    "# split a univariate dataset into train/test sets\n",
    "def train_test_split(data, n_test):\n",
    "\treturn data[:-n_test, :], data[-n_test:, :]\n",
    "\n",
    "# fit an xgboost model and make a one step prediction\n",
    "def xgboost_forecast(train, testX):\n",
    "\t# transform list into array\n",
    "\ttrain = asarray(train)\n",
    "\t# split into input and output columns\n",
    "\ttrainX, trainy = train[:, :-20], train[:, -20:]\n",
    "\t# fit model\n",
    "\t#model = XGBRegressor(objective='reg:squarederror', n_estimators=1000)\n",
    "\t#print(train.shape)\n",
    "\t#print(trainX.shape)\n",
    "\t#print(trainy.shape)\n",
    "    # fitting\n",
    "\t#model = MultiOutputRegressor(xgb.XGBRegressor(objective='reg:squarederror', n_estimators=1000)).fit(trainX, trainy)\n",
    "\tmodel = MultiOutputRegressor(xgb.XGBRegressor(objective='reg:squarederror', n_estimators=100, max_depth=4, eta=0.1, \n",
    "\tmin_child_weight=1,subsample=0.9, colsample_bytree=0.9)).fit(trainX, trainy)\n",
    "    # predicting\n",
    "\n",
    "    #print(np.mean((multioutputregressor.predict(X) - y)**2, axis=0))\n",
    "\t#model.fit(trainX, trainy)\n",
    "\t# make a one-step prediction\n",
    "\t#print(testX)\n",
    "\t#model = MultiOutputRegressor(model).fit(trainX, trainy)\n",
    "\tyhat = model.predict(asarray([testX]))\n",
    "\t#print(\"yhat is :\",yhat)\n",
    "\treturn yhat[0]\n",
    "\n",
    "# walk-forward validation for univariate data\n",
    "def walk_forward_validation(data, n_test):\n",
    "\tpredictions = list()\n",
    "\t# split dataset\n",
    "\ttrain, test = train_test_split(data, n_test)\n",
    "\t#print(test)\n",
    "\t#print(\"train shape is :\",train.shape)\n",
    "\t#print(\"test shape is :\",test.shape)\n",
    "\t# seed history with training dataset\n",
    "\thistory = [x for x in train]\n",
    "\t#print(\"history is \", history)\n",
    "\t# step over each time-step in the test set\n",
    "\tfor i in range(0,len(test),20):\n",
    "\t\t# split test row into input and output columns\n",
    "\t\ttestX, testy = test[i, :-20], test[i, -20:]\n",
    "\t\t#print(\"testX shape is:\", testX.shape)\n",
    "\t\t#print(\"testy shape is :\", testy.shape)\n",
    "\t\t# fit model on history and make a prediction\n",
    "\t\tyhat = xgboost_forecast(history, testX)\n",
    "\t\t# store forecast in list of predictions\n",
    "\t\tpredictions.append(yhat)\n",
    "\t\t# add actual observation to history for the next loop\n",
    "\t\thistory.append(test[i])\n",
    "\t\t# summarize progress\n",
    "\t\t#print('>expected=%.1f, predicted=%.1f' % (testy, yhat))\n",
    "\t\t#print('>expected = ',testy)\n",
    "\t\t#print('>predicted = ',yhat)       \n",
    "\t# estimate prediction error\n",
    "\t#print('>expected = ',testy)\n",
    "\tpredictions2 = [num for sublist in predictions for num in sublist]\n",
    "\t#print(len(predictions2))\n",
    "\t#print(test[:, -2:])\n",
    "\terror = mean_absolute_error(test[:, -1], predictions2)\n",
    "\treturn error, test[:, -1], predictions2\n",
    "\n",
    "\n",
    "# transform the time series data into supervised learning\n",
    "data = series_to_supervised(dataset['TEMP'],60,20)\n",
    "print(data.shape)\n",
    "# evaluate\n",
    "mae, y, yhat = walk_forward_validation(data, 100)\n",
    "#print(y)\n",
    "#print(yhat)\n",
    "print('MAE: %.3f' % mae)\n",
    "# plot expected vs preducted\n",
    "pyplot.plot(y, label='Expected')\n",
    "pyplot.plot(yhat, label='Predicted')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n",
    "\n",
    "mse = mean_squared_error(y, yhat)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(y, yhat)\n",
    "mape = np.mean(np.abs((y - yhat) / y) * 1)\n",
    "print('MSE: {:.6f}'.format(mse))\n",
    "print('RMSE: {:.6f}'.format(rmse))\n",
    "print('MAE: {:.6f}'.format(mae))\n",
    "print('MAPE: {:.6f}%'.format(mape))\n",
    "def smape(actual, forecast):\n",
    "    \"\"\"Symmetric mean absolute percentage error\"\"\"\n",
    "    return np.mean((np.abs(actual - forecast) * 2 / (np.abs(actual) + np.abs(forecast))))\n",
    "smape_score = smape(y, yhat)\n",
    "print(\"SMAPE:{:.6f}\" .format(smape_score))\n",
    "# Calculate R-squared score\n",
    "r2 = r2_score(y,yhat)\n",
    "print(\"R-squared:{:.6f}\" .format(r2))\n",
    "#plot forecasts against actual outcomes\n",
    "pyplot.plot(y)\n",
    "pyplot.plot(yhat, color='red')\n",
    "pyplot.show()\n",
    "\n",
    "end = time.time()\n",
    "print(end - start)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8feeb16",
   "metadata": {},
   "source": [
    "LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "id": "134dd447",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: total: 0 ns\n",
      "Wall time: 0 ns\n",
      "0.007214999990537763\n",
      "(3142, 9)\n",
      "[26.75  26.75  32.    32.    25.375 25.375 25.375 25.375]\n",
      "(3132, 1)\n",
      "(3132, 8)\n",
      "WARNING:tensorflow:Layer lstm_150 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_150_input'), name='lstm_150_input', description=\"created by layer 'lstm_150_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_150_input'), name='lstm_150_input', description=\"created by layer 'lstm_150_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "196/196 [==============================] - 4s 15ms/step - loss: 49.4709\n",
      "Epoch 2/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 10.0074\n",
      "Epoch 3/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 8.3535\n",
      "Epoch 4/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 7.5498\n",
      "Epoch 5/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 7.7269\n",
      "Epoch 6/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 7.2772\n",
      "Epoch 7/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 7.0514\n",
      "Epoch 8/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 6.8234\n",
      "Epoch 9/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 7.3210\n",
      "Epoch 10/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 7.0063\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_150_input'), name='lstm_150_input', description=\"created by layer 'lstm_150_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=25.1, predicted=25.8\n",
      "[26.75       32.         32.         25.375      25.375      25.375\n",
      " 25.375      25.06200027]\n",
      "(3133, 1)\n",
      "(3133, 8)\n",
      "WARNING:tensorflow:Layer lstm_151 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_151_input'), name='lstm_151_input', description=\"created by layer 'lstm_151_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_151_input'), name='lstm_151_input', description=\"created by layer 'lstm_151_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "196/196 [==============================] - 4s 15ms/step - loss: 34.2273\n",
      "Epoch 2/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 10.6273\n",
      "Epoch 3/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 9.5597\n",
      "Epoch 4/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 9.1432\n",
      "Epoch 5/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 9.2345\n",
      "Epoch 6/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.9475\n",
      "Epoch 7/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 7.7629\n",
      "Epoch 8/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.1376\n",
      "Epoch 9/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 7.3447\n",
      "Epoch 10/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 7.2042\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_151_input'), name='lstm_151_input', description=\"created by layer 'lstm_151_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=25.1, predicted=25.8\n",
      "[32.         32.         25.375      25.375      25.375      25.375\n",
      " 25.06200027 25.06200027]\n",
      "(3134, 1)\n",
      "(3134, 8)\n",
      "WARNING:tensorflow:Layer lstm_152 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_152_input'), name='lstm_152_input', description=\"created by layer 'lstm_152_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_152_input'), name='lstm_152_input', description=\"created by layer 'lstm_152_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "196/196 [==============================] - 4s 15ms/step - loss: 41.6470\n",
      "Epoch 2/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 12.0668\n",
      "Epoch 3/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 9.2211\n",
      "Epoch 4/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 9.2901\n",
      "Epoch 5/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 8.8793\n",
      "Epoch 6/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.3391\n",
      "Epoch 7/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.5220\n",
      "Epoch 8/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 8.7400\n",
      "Epoch 9/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.5803\n",
      "Epoch 10/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.5667\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_152_input'), name='lstm_152_input', description=\"created by layer 'lstm_152_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=25.1, predicted=26.0\n",
      "[32.         25.375      25.375      25.375      25.375      25.06200027\n",
      " 25.06200027 25.06200027]\n",
      "(3135, 1)\n",
      "(3135, 8)\n",
      "WARNING:tensorflow:Layer lstm_153 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_153_input'), name='lstm_153_input', description=\"created by layer 'lstm_153_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_153_input'), name='lstm_153_input', description=\"created by layer 'lstm_153_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "196/196 [==============================] - 4s 16ms/step - loss: 62.6601\n",
      "Epoch 2/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 10.5038\n",
      "Epoch 3/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 9.4572\n",
      "Epoch 4/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 9.3036\n",
      "Epoch 5/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 9.0583\n",
      "Epoch 6/10\n",
      "196/196 [==============================] - 3s 14ms/step - loss: 8.8926\n",
      "Epoch 7/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 9.0491\n",
      "Epoch 8/10\n",
      "196/196 [==============================] - 3s 16ms/step - loss: 8.8535\n",
      "Epoch 9/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 9.0577\n",
      "Epoch 10/10\n",
      "196/196 [==============================] - 3s 15ms/step - loss: 8.6202\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_153_input'), name='lstm_153_input', description=\"created by layer 'lstm_153_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">expected=25.1, predicted=25.6\n",
      "[25.375      25.375      25.375      25.375      25.06200027 25.06200027\n",
      " 25.06200027 25.06200027]\n",
      "(3136, 1)\n",
      "(3136, 8)\n",
      "WARNING:tensorflow:Layer lstm_154 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_154_input'), name='lstm_154_input', description=\"created by layer 'lstm_154_input'\"), but it was called on an input with incompatible shape (16, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_154_input'), name='lstm_154_input', description=\"created by layer 'lstm_154_input'\"), but it was called on an input with incompatible shape (16, 8, 1).\n",
      "196/196 [==============================] - 2s 8ms/step - loss: 64.4750\n",
      "Epoch 2/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 9.1397\n",
      "Epoch 3/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 8.3232\n",
      "Epoch 4/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 7.8808\n",
      "Epoch 5/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 8.0133\n",
      "Epoch 6/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 7.7437\n",
      "Epoch 7/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 7.1293\n",
      "Epoch 8/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 7.1566\n",
      "Epoch 9/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 7.1749\n",
      "Epoch 10/10\n",
      "196/196 [==============================] - 1s 7ms/step - loss: 6.8425\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_154_input'), name='lstm_154_input', description=\"created by layer 'lstm_154_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=25.1, predicted=24.0\n",
      "[25.375      25.375      25.375      25.06200027 25.06200027 25.06200027\n",
      " 25.06200027 25.06200027]\n",
      "(3137, 1)\n",
      "(3137, 8)\n",
      "WARNING:tensorflow:Layer lstm_155 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_155_input'), name='lstm_155_input', description=\"created by layer 'lstm_155_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_155_input'), name='lstm_155_input', description=\"created by layer 'lstm_155_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "197/197 [==============================] - 4s 14ms/step - loss: 59.7449\n",
      "Epoch 2/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.6558\n",
      "Epoch 3/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.1553\n",
      "Epoch 4/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 7.9116\n",
      "Epoch 5/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.1827\n",
      "Epoch 6/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.3151\n",
      "Epoch 7/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.3651\n",
      "Epoch 8/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.1332\n",
      "Epoch 9/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.0498\n",
      "Epoch 10/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 6.9628\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_155_input'), name='lstm_155_input', description=\"created by layer 'lstm_155_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=25.1, predicted=24.8\n",
      "[25.375      25.375      25.06200027 25.06200027 25.06200027 25.06200027\n",
      " 25.06200027 25.06200027]\n",
      "(3138, 1)\n",
      "(3138, 8)\n",
      "WARNING:tensorflow:Layer lstm_156 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_156_input'), name='lstm_156_input', description=\"created by layer 'lstm_156_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_156_input'), name='lstm_156_input', description=\"created by layer 'lstm_156_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "197/197 [==============================] - 4s 15ms/step - loss: 27.6079\n",
      "Epoch 2/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 11.0633\n",
      "Epoch 3/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.6622\n",
      "Epoch 4/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.3890\n",
      "Epoch 5/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 9.0309\n",
      "Epoch 6/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.9163\n",
      "Epoch 7/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 8.6876\n",
      "Epoch 8/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 8.3497\n",
      "Epoch 9/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 9.1113\n",
      "Epoch 10/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 8.7626\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_156_input'), name='lstm_156_input', description=\"created by layer 'lstm_156_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=32.0, predicted=24.9\n",
      "[25.375      25.06200027 25.06200027 25.06200027 25.06200027 25.06200027\n",
      " 25.06200027 32.        ]\n",
      "(3139, 1)\n",
      "(3139, 8)\n",
      "WARNING:tensorflow:Layer lstm_157 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_157_input'), name='lstm_157_input', description=\"created by layer 'lstm_157_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_157_input'), name='lstm_157_input', description=\"created by layer 'lstm_157_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "197/197 [==============================] - 4s 15ms/step - loss: 52.9624\n",
      "Epoch 2/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 10.2017\n",
      "Epoch 3/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 10.2082\n",
      "Epoch 4/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 10.0100\n",
      "Epoch 5/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 9.0891\n",
      "Epoch 6/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.1115\n",
      "Epoch 7/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.9330\n",
      "Epoch 8/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.2559\n",
      "Epoch 9/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.8819\n",
      "Epoch 10/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.1699\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_157_input'), name='lstm_157_input', description=\"created by layer 'lstm_157_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">expected=32.0, predicted=27.0\n",
      "[25.06200027 25.06200027 25.06200027 25.06200027 25.06200027 25.06200027\n",
      " 32.         32.        ]\n",
      "(3140, 1)\n",
      "(3140, 8)\n",
      "WARNING:tensorflow:Layer lstm_158 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_158_input'), name='lstm_158_input', description=\"created by layer 'lstm_158_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_158_input'), name='lstm_158_input', description=\"created by layer 'lstm_158_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "197/197 [==============================] - 4s 15ms/step - loss: 44.0963\n",
      "Epoch 2/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.6738\n",
      "Epoch 3/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.3298\n",
      "Epoch 4/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.4767\n",
      "Epoch 5/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.5791\n",
      "Epoch 6/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.6158\n",
      "Epoch 7/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.5921\n",
      "Epoch 8/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.1619\n",
      "Epoch 9/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.0489\n",
      "Epoch 10/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.1876\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_158_input'), name='lstm_158_input', description=\"created by layer 'lstm_158_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=23.8, predicted=27.2\n",
      "[25.06200027 25.06200027 25.06200027 25.06200027 25.06200027 32.\n",
      " 32.         23.75      ]\n",
      "(3141, 1)\n",
      "(3141, 8)\n",
      "WARNING:tensorflow:Layer lstm_159 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_159_input'), name='lstm_159_input', description=\"created by layer 'lstm_159_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_159_input'), name='lstm_159_input', description=\"created by layer 'lstm_159_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      "197/197 [==============================] - 4s 15ms/step - loss: 34.3465\n",
      "Epoch 2/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 9.5116\n",
      "Epoch 3/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.9601\n",
      "Epoch 4/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 8.2149\n",
      "Epoch 5/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.5389\n",
      "Epoch 6/10\n",
      "197/197 [==============================] - 3s 14ms/step - loss: 7.7917\n",
      "Epoch 7/10\n",
      "197/197 [==============================] - 3s 16ms/step - loss: 6.9334\n",
      "Epoch 8/10\n",
      "197/197 [==============================] - 3s 16ms/step - loss: 7.2612\n",
      "Epoch 9/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 7.0193\n",
      "Epoch 10/10\n",
      "197/197 [==============================] - 3s 15ms/step - loss: 6.6608\n",
      "WARNING:tensorflow:Model was constructed with shape (None, 2, 1) for input KerasTensor(type_spec=TensorSpec(shape=(None, 2, 1), dtype=tf.float32, name='lstm_159_input'), name='lstm_159_input', description=\"created by layer 'lstm_159_input'\"), but it was called on an input with incompatible shape (None, 8, 1).\n",
      ">expected=32.0, predicted=27.9\n",
      "MAE: 2.391\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 10.676758\n",
      "RMSE: 3.267531\n",
      "MAE: 2.391116\n",
      "MAPE: 9.352086%\n",
      "SMAPE:9.851091\n",
      "R-squared:0.012290\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: total: 0 ns\n",
      "Wall time: 0 ns\n"
     ]
    }
   ],
   "source": [
    "%time \n",
    "import timeit\n",
    "%timeit\n",
    "print(timeit.timeit())\n",
    "\n",
    "# forecast monthly births with xgboost\n",
    "from numpy import asarray\n",
    "from pandas import read_csv\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from xgboost import XGBRegressor\n",
    "from matplotlib import pyplot\n",
    "\n",
    "# Importing necessary libraries\n",
    "import numpy as np\n",
    "from keras.models import Sequential\n",
    "from keras.layers import LSTM, Dense, Dropout\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# transform a time series dataset into a supervised learning dataset\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\tn_vars = 1 if type(data) is list else data.shape[0]\n",
    "\tdf = DataFrame(data)\n",
    "\tcols = list()\n",
    "\t# input sequence (t-n, ... t-1)\n",
    "\tfor i in range(n_in, 0, -1):\n",
    "\t\tcols.append(df.shift(i))\n",
    "\t# forecast sequence (t, t+1, ... t+n)\n",
    "\tfor i in range(0, n_out):\n",
    "\t\tcols.append(df.shift(-i))\n",
    "\t# put it all together\n",
    "\tagg = concat(cols, axis=1)\n",
    "\t# drop rows with NaN values\n",
    "\tif dropnan:\n",
    "\t\tagg.dropna(inplace=True)\n",
    "\treturn agg.values\n",
    "\n",
    "# split a univariate dataset into train/test sets\n",
    "def train_test_split(data, n_test):\n",
    "\treturn data[:-n_test, :], data[-n_test:, :]\n",
    "\n",
    "# fit an xgboost model and make a one step prediction\n",
    "def lstm_forecast(train, testX):\n",
    "    # transform list into array\n",
    "    train = asarray(train)\n",
    "    # split into input and output columns\n",
    "    trainX, trainy = train[:, :-1], train[:, -1]\n",
    "    # fit model\n",
    "    trainy.shape = (trainy.shape[0], 1)\n",
    "    print(testX)\n",
    "    print(trainy.shape)\n",
    "    print(trainX.shape)\n",
    "       # Reshaping the input data to fit the LSTM network\n",
    "    trainX = trainX.reshape((trainX.shape[0], trainX.shape[1],1))\n",
    "    trainy = trainy.reshape((trainy.shape[0], trainy.shape[1], 1))\n",
    "\n",
    "        # Creating the LSTM model\n",
    "    model = Sequential()\n",
    "    model.add(LSTM(64, activation='relu', input_shape=(2, 1)))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(1))\n",
    "    \n",
    "    # Build the LSTM model\n",
    "    #model = Sequential()\n",
    "    #model.add(LSTM(128, activation='relu', input_shape=(2, 1)))\n",
    "    #model.add(Dropout(0.2))\n",
    "    #model.add(Dense(64, activation='relu'))\n",
    "    #model.add(Dropout(0.2))\n",
    "    #model.add(Dense(32, activation='relu'))\n",
    "    #model.add(Dropout(0.2))\n",
    "    #model.add(Dense(1))\n",
    "    \n",
    "    model.compile(optimizer='adam', loss='mse')\n",
    "\n",
    "        # Training the model\n",
    "    model.fit(trainX, trainy, epochs=10, batch_size=16, verbose=1 )\n",
    "\n",
    "\n",
    "    #model.fit(trainX, trainy)\n",
    "        #multi-step prediction\n",
    "    #rained_model = MultiOutputRegressor(model).fit(trainX, trainy)\n",
    "        #train_forecasts = trained_model.predict(x_train)\n",
    "    # make a one-step prediction\n",
    "    #print(testX)\n",
    "    testX=np.asarray(testX)\n",
    "    testX.shape = (testX.shape[0], 1)\n",
    "    \n",
    "    yhat = model.predict(asarray([testX]))\n",
    "    #print(yhat)\n",
    "    return yhat[0]\n",
    "\n",
    "# walk-forward validation for univariate data\n",
    "def walk_forward_validation(data, n_test):\n",
    "    predictions = list()\n",
    "    # split dataset\n",
    "    train, test = train_test_split(data, n_test)\n",
    "    # seed history with training dataset\n",
    "    history = [x for x in train]\n",
    "    # step over each time-step in the test set\n",
    "    for i in range(len(test)):\n",
    "        # split test row into input and output columns\n",
    "        testX, testy = test[i, :-1], test[i, -1]\n",
    "        # fit model on history and make a prediction\n",
    "        yhat = lstm_forecast(history, testX)\n",
    "        # store forecast in list of predictions\n",
    "        predictions.append(yhat)\n",
    "        # add actual observation to history for the next loop\n",
    "        history.append(test[i])\n",
    "        # summarize progress\n",
    "        print('>expected=%.1f, predicted=%.1f' % (testy, yhat))\n",
    "    # estimate prediction error\n",
    "    error = mean_absolute_error(test[:, -1], predictions)\n",
    "    return error, test[:, -1], predictions\n",
    "\n",
    "\n",
    "# transform the time series data into supervised learning\n",
    "data = series_to_supervised(dataset['TEMP'], n_in=8)\n",
    "print(data.shape)\n",
    "# evaluate\n",
    "mae, y, yhat = walk_forward_validation(data, 10)\n",
    "print('MAE: %.3f' % mae)\n",
    "# plot expected vs preducted\n",
    "pyplot.plot(y, label='Expected')\n",
    "pyplot.plot(yhat, label='Predicted')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n",
    "\n",
    "\n",
    "mse = mean_squared_error(y, yhat)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(y, yhat)\n",
    "mape = np.mean(np.abs((y - yhat) / y) * 100)\n",
    "print('MSE: {:.6f}'.format(mse))\n",
    "print('RMSE: {:.6f}'.format(rmse))\n",
    "print('MAE: {:.6f}'.format(mae))\n",
    "print('MAPE: {:.6f}%'.format(mape))\n",
    "def smape(actual, forecast):\n",
    "    \"\"\"Symmetric mean absolute percentage error\"\"\"\n",
    "    return np.mean((np.abs(actual - forecast) * 200 / (np.abs(actual) + np.abs(forecast))))\n",
    "smape_score = smape(y, yhat)\n",
    "print(\"SMAPE:{:.6f}\" .format(smape_score))\n",
    "# Calculate R-squared score\n",
    "r2 = r2_score(y,yhat)\n",
    "print(\"R-squared:{:.6f}\" .format(r2))\n",
    "#plot forecasts against actual outcomes\n",
    "pyplot.plot(y)\n",
    "pyplot.plot(yhat, color='red')\n",
    "pyplot.show()\n",
    "%time "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "e388e029",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3129, 22)\n",
      "(3089, 20)\n",
      "(3089, 2)\n",
      "WARNING:tensorflow:Layer lstm_39 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "258/258 [==============================] - 4s 11ms/step - loss: 114.2543\n",
      "Epoch 2/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 13.4493\n",
      "Epoch 3/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 9.4780\n",
      "Epoch 4/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 8.1021\n",
      "Epoch 5/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 7.3896\n",
      "Epoch 6/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 6.5798\n",
      "Epoch 7/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 6.1634\n",
      "Epoch 8/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 6.0640\n",
      "Epoch 9/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 5.5954\n",
      "Epoch 10/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 5.5781\n",
      ">expected =  [26.75 26.75 26.75 26.75 26.75 26.75 26.75 26.75 26.75 28.75 28.75 28.75\n",
      " 28.75 28.75 28.75 28.75 28.75 28.75 28.75 28.75]\n",
      ">predicted =  [26.020918 26.043905 25.991318 26.035587 26.020693 25.994919 26.05935\n",
      " 26.039507 26.017544 25.963629 26.03898  26.054798 25.992823 25.962086\n",
      " 26.025524 25.925125 25.949648 25.919819 25.964314 25.953674]\n",
      "(3090, 20)\n",
      "(3090, 2)\n",
      "WARNING:tensorflow:Layer lstm_40 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Epoch 1/10\n",
      "258/258 [==============================] - 4s 11ms/step - loss: 120.2795\n",
      "Epoch 2/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 14.2673\n",
      "Epoch 3/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 10.7456\n",
      "Epoch 4/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 9.4188\n",
      "Epoch 5/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 8.1297\n",
      "Epoch 6/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 7.8548\n",
      "Epoch 7/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 7.1193\n",
      "Epoch 8/10\n",
      "258/258 [==============================] - 3s 11ms/step - loss: 6.7431\n",
      "Epoch 9/10\n",
      "258/258 [==============================] - 3s 12ms/step - loss: 6.8998\n",
      "Epoch 10/10\n",
      "258/258 [==============================] - 3s 12ms/step - loss: 6.6033\n",
      ">expected =  [28.75 28.75 28.75 28.75 28.75 28.75 28.75 28.75 28.75 28.75 28.75 28.75\n",
      " 28.75 28.75 28.75 26.75 26.75 26.75 26.75 26.75]\n",
      ">predicted =  [28.93037  28.865532 28.905027 28.919903 28.880297 28.896442 28.933012\n",
      " 28.85108  28.91844  28.872059 28.89509  28.859514 28.850473 28.805532\n",
      " 28.798788 28.799885 28.795345 28.787388 28.745264 28.807636]\n",
      "40\n",
      "[[28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       28.75      ]\n",
      " [28.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       26.75      ]\n",
      " [26.75       32.        ]\n",
      " [32.         32.        ]\n",
      " [32.         25.375     ]\n",
      " [25.375      25.375     ]\n",
      " [25.375      25.375     ]\n",
      " [25.375      25.375     ]\n",
      " [25.375      25.06200027]\n",
      " [25.06200027 25.06200027]\n",
      " [25.06200027 25.06200027]\n",
      " [25.06200027 25.06200027]\n",
      " [25.06200027 25.06200027]\n",
      " [25.06200027 25.06200027]\n",
      " [25.06200027 32.        ]\n",
      " [32.         32.        ]\n",
      " [32.         23.75      ]\n",
      " [23.75       32.        ]]\n",
      "MAE: 2.830\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%time \n",
    "import timeit\n",
    "%timeit\n",
    "print(timeit.timeit())\n",
    "\n",
    "# forecast monthly births with xgboost\n",
    "from numpy import asarray\n",
    "from pandas import read_csv\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from xgboost import XGBRegressor\n",
    "from matplotlib import pyplot\n",
    "\n",
    "# Importing necessary libraries\n",
    "import numpy as np\n",
    "from keras.models import Sequential\n",
    "from keras.layers import LSTM, Dense, Dropout\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# transform a time series dataset into a supervised learning dataset\n",
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "\tn_vars = 1 if type(data) is list else data.shape[0]\n",
    "\tdf = DataFrame(data)\n",
    "\tcols = list()\n",
    "\t# input sequence (t-n, ... t-1)\n",
    "\tfor i in range(n_in, 0, -1):\n",
    "\t\tcols.append(df.shift(i))\n",
    "\t# forecast sequence (t, t+1, ... t+n)\n",
    "\tfor i in range(0, n_out):\n",
    "\t\tcols.append(df.shift(-i))\n",
    "\t# put it all together\n",
    "\tagg = concat(cols, axis=1)\n",
    "\t# drop rows with NaN values\n",
    "\tif dropnan:\n",
    "\t\tagg.dropna(inplace=True)\n",
    "\treturn agg.values\n",
    "\n",
    "# split a univariate dataset into train/test sets\n",
    "def train_test_split(data, n_test):\n",
    "\treturn data[:-n_test, :], data[-n_test:, :]\n",
    "\n",
    "# fit an xgboost model and make a one step prediction\n",
    "def lstm_forecast(train, testX):\n",
    "    # transform list into array\n",
    "    train = asarray(train)\n",
    "    # split into input and output columns\n",
    "    trainX, trainy = train[:, :-20], train[:, -20:]\n",
    "    # fit model\n",
    "    #trainy.shape = (trainy.shape[0], 1)\n",
    "    #print(testX)\n",
    "    print(trainy.shape)\n",
    "    print(trainX.shape)\n",
    "       # Reshaping the input data to fit the LSTM network\n",
    "    trainX = trainX.reshape((trainX.shape[0], trainX.shape[1],1))\n",
    "    trainy = trainy.reshape((trainy.shape[0], trainy.shape[1], 1))\n",
    "\n",
    "        # Creating the LSTM model\n",
    "    #model = Sequential()\n",
    "    #model.add(LSTM(64, activation='relu', input_shape=(2, 1)))\n",
    "    #model.add(Dropout(0.2))\n",
    "    #model.add(Dense(20))\n",
    "    \n",
    "\n",
    "    model = Sequential()\n",
    "    model.add(LSTM(128, activation='relu', input_shape=(2, 1)))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(64, activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(32, activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(20))\n",
    "    \n",
    "    model.compile(optimizer='adam', loss='mse')\n",
    "\n",
    "        # Training the model\n",
    "    model.fit(trainX, trainy, epochs=10, batch_size=12, verbose=1 )\n",
    "\n",
    "2\n",
    "    #print(testX)\n",
    "    testX=np.asarray(testX)\n",
    "    testX.shape = (testX.shape[0], 1)\n",
    "    \n",
    "    yhat = model.predict(asarray([testX]))\n",
    "    #print(yhat)\n",
    "    return yhat[0]\n",
    "\n",
    "# walk-forward validation for univariate data   \n",
    "def walk_forward_validation(data, n_test):\n",
    "    predictions = list()\n",
    "    # split dataset\n",
    "    train, test = train_test_split(data, n_test)\n",
    "    # seed history with training dataset\n",
    "    history = [x for x in train]\n",
    "    # step over each time-step in the test set\n",
    "    for i in range(0, len(test), 20):\n",
    "        # split test row into input and output columns\n",
    "        testX, testy = test[i, :-20], test[i, -20:]\n",
    "        # fit model on history and make a prediction\n",
    "        yhat = lstm_forecast(history, testX)\n",
    "        # store forecast in list of predictions\n",
    "        predictions.append(yhat)\n",
    "        # add actual observation to history for the next loop\n",
    "        history.append(test[i])\n",
    "        # summarize progress\n",
    "        #print('>expected=%.1f, predicted=%.1f' % (testy, yhat))\n",
    "        print('>expected = ',testy)\n",
    "        print('>predicted = ',yhat)\n",
    "    # estimate prediction error\n",
    "    predictions2 = [num for sublist in predictions for num in sublist]\n",
    "    print(len(predictions2))\n",
    "    print(test[:, -2:])\n",
    "    error = mean_absolute_error(test[:, -1], predictions2)\n",
    "    return error, test[:, -1], predictions2\n",
    "\n",
    "    #error = mean_absolute_error(test[:, -1], predictions)\n",
    "    #return error, test[:, -1], predictions\n",
    "\n",
    "# transform the time series data into supervised learning\n",
    "data = series_to_supervised(dataset['TEMP'], 2,20)\n",
    "print(data.shape)\n",
    "# evaluate\n",
    "mae, y, yhat = walk_forward_validation(data, 40)\n",
    "print('MAE: %.3f' % mae)\n",
    "# plot expected vs preducted\n",
    "pyplot.plot(y, label='Expected')\n",
    "pyplot.plot(yhat, label='Predicted')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n",
    "\n",
    "end = time.time()\n",
    "print(end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "47c00cf5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Layer lstm_64 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<PIL.Image.Image image mode=RGBA size=555x63>"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import visualkeras\n",
    "#model = Sequential()\n",
    "#model.add(LSTM(64, activation='relu', input_shape=(2, 1)))\n",
    "#model.add(Dropout(0.2))\n",
    "#model.add(Dense(20))\n",
    "\n",
    "# Build the LSTM model\n",
    "model = Sequential()\n",
    "model.add(LSTM(128, activation='relu', input_shape=(2, 1)))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(64, activation='relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(32, activation='relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(1))\n",
    "\n",
    "\n",
    "#visualkeras.graph_view(model,end=).show()\n",
    "visualkeras.layered_view(model, legend=True,scale_xy=1, scale_z=1, max_z=1000)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "650b95d6",
   "metadata": {},
   "source": [
    "ARIMA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "id": "8ca72981",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\aggelosuser\\AppData\\Local\\Temp\\ipykernel_36432\\571919582.py:6: FutureWarning: The pandas.datetime class is deprecated and will be removed from pandas in a future version. Import from datetime module instead.\n",
      "  from pandas import datetime\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE: 16.544\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 273.711216\n",
      "RMSE: 16.544220\n",
      "MAE: 9.899139\n",
      "MAPE: 0.809176%\n",
      "SMAPE:0.596940\n",
      "R-squared:-0.025250\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18.265939712524414\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.time()\n",
    "\n",
    "# evaluate an ARIMA model using a walk-forward validation\n",
    "from pandas import read_csv\n",
    "from pandas import datetime\n",
    "from matplotlib import pyplot\n",
    "from statsmodels.tsa.arima.model import ARIMA\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from math import sqrt\n",
    "# load dataset\n",
    "series = dataset['CPU_USAGE']\n",
    "\n",
    "# split into train and test sets\n",
    "X = series.values\n",
    "size = 100 # int(len(X) * 0.66)\n",
    "train, test = X[:-size], X[-size:]\n",
    "history = [x for x in train]\n",
    "predictions = list()\n",
    "# walk-forward validation\n",
    "for t in range(len(test)):\n",
    "    model = ARIMA(history, order=(0,0,1))\n",
    "    model_fit = model.fit()\n",
    "    output = model_fit.forecast()\n",
    "    yhat = output[0]\n",
    "    predictions.append(yhat)\n",
    "    obs = test[t]\n",
    "    history.append(obs)\n",
    "#    print('predicted=%f, expected=%f' % (yhat, obs))\n",
    "# evaluate forecasts\n",
    "rmse = sqrt(mean_squared_error(test, predictions))\n",
    "print('Test RMSE: %.3f' % rmse)\n",
    "# plot forecasts against actual outcomes\n",
    "pyplot.plot(test)\n",
    "pyplot.plot(predictions, color='red')\n",
    "pyplot.show()\n",
    "\n",
    "mse = mean_squared_error(test[:len(predictions)], predictions)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(test[:len(predictions)], predictions)\n",
    "mape = np.mean(np.abs((test[:len(predictions)] - predictions) / test[:len(predictions)])) * 1\n",
    "print('MSE: {:.6f}'.format(mse))\n",
    "print('RMSE: {:.6f}'.format(rmse))\n",
    "print('MAE: {:.6f}'.format(mae))\n",
    "print('MAPE: {:.6f}%'.format(mape))\n",
    "def smape(actual, forecast):\n",
    "    \"\"\"Symmetric mean absolute percentage error\"\"\"\n",
    "    return np.mean((np.abs(actual - forecast) * 2 / (np.abs(actual) + np.abs(forecast))))\n",
    "smape_score = smape(test[:len(predictions)], predictions)\n",
    "print(\"SMAPE:{:.6f}\" .format(smape_score))\n",
    "# Calculate R-squared score\n",
    "r2 = r2_score(test[:len(predictions)],predictions)\n",
    "print(\"R-squared:{:.6f}\" .format(r2))\n",
    "#plot forecasts against actual outcomes\n",
    "pyplot.plot(test)\n",
    "pyplot.plot(predictions, color='red')\n",
    "pyplot.show()\n",
    "\n",
    "end = time.time()\n",
    "print(end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "id": "9ee05484",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\aggelosuser\\AppData\\Local\\Temp\\ipykernel_36432\\90526236.py:5: FutureWarning: The pandas.datetime class is deprecated and will be removed from pandas in a future version. Import from datetime module instead.\n",
      "  from pandas import datetime\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 273.466530\n",
      "RMSE: 16.536823\n",
      "MAE: 9.897984\n",
      "MAPE: 0.809033%\n",
      "SMAPE:0.596938\n",
      "R-squared:-0.024334\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "57.489888429641724\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.time()\n",
    "\n",
    "from pandas import read_csv\n",
    "from pandas import datetime\n",
    "from matplotlib import pyplot\n",
    "from statsmodels.tsa.arima.model import ARIMA\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from math import sqrt\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "# load dataset\n",
    "series = dataset['CPU_USAGE']\n",
    "\n",
    "# split into train and test sets\n",
    "X = series.values\n",
    "#size = int(len(X) * 0.66)\n",
    "size = 100 # int(len(X) * 0.66)\n",
    "train, test = X[:-size], X[-size:]\n",
    "#train, test = X[0:size], X[size:len(X)]\n",
    "history = [x for x in train]\n",
    "predictions = list()\n",
    "horizon = 1  # set the forecasting horizon to 5\n",
    "\n",
    "# iterate over the test set and predict 5 steps at a time\n",
    "for t in range(0, len(test), horizon):\n",
    "    if t+horizon > len(test):\n",
    "        break  # if less than 5 samples left in the test set, break out of the loop\n",
    "    model = ARIMA(history, order=(1,1,1))\n",
    "    model_fit = model.fit()\n",
    "    output = model_fit.forecast(steps=horizon)\n",
    "    #print(output)\n",
    "    #yhat = output[0]\n",
    "    predictions.extend(output.tolist())\n",
    "    history.extend(test[t:t+horizon])\n",
    "    # print('predicted=%s, expected=%s' % (yhat, test[t:t+horizon]))\n",
    "\n",
    "# evaluate forecasts\n",
    "#rmse = sqrt(mean_squared_error(test[:len(predictions)], predictions))\n",
    "#print('Test RMSE: %.3f' % rmse)\n",
    "# Step 5: Evaluate the performance of the ARIMA model\n",
    "mse = mean_squared_error(test[:len(predictions)], predictions)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(test[:len(predictions)], predictions)\n",
    "mape = np.mean(np.abs((test[:len(predictions)] - predictions) / test[:len(predictions)])) * 1\n",
    "print('MSE: {:.6f}'.format(mse))\n",
    "print('RMSE: {:.6f}'.format(rmse))\n",
    "print('MAE: {:.6f}'.format(mae))\n",
    "print('MAPE: {:.6f}%'.format(mape))\n",
    "def smape(actual, forecast):\n",
    "    \"\"\"Symmetric mean absolute percentage error\"\"\"\n",
    "    return np.mean((np.abs(actual - forecast) * 2 / (np.abs(actual) + np.abs(forecast))))\n",
    "smape_score = smape(test[:len(predictions)], predictions)\n",
    "print(\"SMAPE:{:.6f}\" .format(smape_score))\n",
    "# Calculate R-squared score\n",
    "r2 = r2_score(test[:len(predictions)],predictions)\n",
    "print(\"R-squared:{:.6f}\" .format(r2))\n",
    "#plot forecasts against actual outcomes\n",
    "pyplot.plot(test)\n",
    "pyplot.plot(predictions, color='red')\n",
    "pyplot.show()\n",
    "\n",
    "end = time.time()\n",
    "print(end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8665074f",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}