MODEL_EVALUATOR.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Library Import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSTM Model Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "350\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEKCAYAAAA4t9PUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIlFKKeURTSRKKaU88v+WypIhWJT11wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2b405f0e438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEKCAYAAADjDHn2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2b4062e4a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEKCAYAAADjDHn2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nNdrH7prkQJoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2b40639ce48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "< OPT. SCORE_SUM EPOCH  247 > : -2.8194185377\n",
      "opt. DEV MSE :  0.178619826455\n",
      "opt. TEST1 MSE :  0.188925073761\n",
      "opt. TEST2 MSE :  0.215351757131\n"
     ]
    }
   ],
   "source": [
    "Reg = False\n",
    "ELBOW = 5\n",
    "d = 'hybrid_LSTM'\n",
    "\n",
    "if Reg :\n",
    "    d += '_with_reg'\n",
    "scores = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/models/\"+d+\".csv\")\n",
    "mse_columns = ['TRAIN_MSE','DEV_MSE']\n",
    "mae_columns = ['TRAIN_MAE','DEV_MAE']\n",
    "test_columns = ['TEST1_MSE', 'TEST2_MSE']\n",
    "\n",
    "#print(scores)\n",
    "\n",
    "end_epoch = scores.shape[0]\n",
    "print(end_epoch)\n",
    "plt.plot(scores[mse_columns[0]][:end_epoch],'--')\n",
    "plt.plot(scores[mse_columns[1]][:end_epoch])\n",
    "plt.legend()\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Mean Squared Error')\n",
    "plt.show()\n",
    "plt.close()\n",
    "plt.plot(scores[mae_columns[0]][:end_epoch],'--')\n",
    "plt.plot(scores[mae_columns[1]][:end_epoch])\n",
    "plt.legend()\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Mean Absolute Error')\n",
    "plt.show()\n",
    "plt.close()\n",
    "plt.plot(scores[test_columns[0]][:end_epoch],'--')\n",
    "plt.plot(scores[test_columns[1]][:end_epoch])\n",
    "plt.legend()\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Mean Squared Error')\n",
    "plt.show()\n",
    "plt.close()\n",
    "\n",
    "\n",
    "\n",
    "score_diff = (scores[mse_columns[1]]-scores[mse_columns[0]])[ELBOW:]\n",
    "\n",
    "score_sum = (scores[mse_columns[1]]+scores[mse_columns[0]])[ELBOW:]\n",
    "\n",
    "score_diff_norm = (score_diff - np.mean(score_diff))/np.std(score_diff)\n",
    "score_sum_norm = (score_sum - np.mean(score_sum))/np.std(score_sum)\n",
    "score_total = score_diff_norm + score_sum_norm\n",
    "idx = np.argmin(score_total)\n",
    "print('< OPT. SCORE_SUM EPOCH ',str(idx+1),'> : '+str(score_total[idx]))\n",
    "print('opt. DEV MSE : ',str(scores[mse_columns[1]][idx]))\n",
    "print('opt. TEST1 MSE : ',str(scores[test_columns[0]][idx]))\n",
    "print('opt. TEST2 MSE : ',str(scores[test_columns[1]][idx]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Other Model Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dev = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/train_dev_test/before_arima/dev.csv\")\n",
    "#dev_Y = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/train_dev_test/after_arima/dev_Y.csv\")\n",
    "test1 = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/train_dev_test/before_arima/test1.csv\")\n",
    "#test1_Y = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/train_dev_test/after_arima/test1_Y.csv\")\n",
    "test2 = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/train_dev_test/before_arima/test2.csv\")\n",
    "#test2_Y = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/train_dev_test/after_arima/test2_Y.csv\")\n",
    "\n",
    "dev = dev.loc[:, ~dev.columns.str.contains('^Unnamed')]\n",
    "#dev_Y = dev_Y.loc[:, ~dev_Y.columns.str.contains('^Unnamed')]\n",
    "test1 = test1.loc[:, ~test1.columns.str.contains('^Unnamed')]\n",
    "#test1_Y = test1_Y.loc[:, ~test1_Y.columns.str.contains('^Unnamed')]\n",
    "test2 = test2.loc[:, ~test2.columns.str.contains('^Unnamed')]\n",
    "#test2_Y = test2_Y.loc[:, ~test2_Y.columns.str.contains('^Unnamed')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Historical Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "STEP = 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.459729626745/0.54492569017\n",
      "0.500474375728/0.574133862615\n",
      "0.445800739647/0.534481179475\n"
     ]
    }
   ],
   "source": [
    "dev_pred = np.array(dev['20'])\n",
    "dev_y = np.array(dev['21']).reshape(1,int(1117500/STEP))[0]\n",
    "test1_pred = np.array(test1['21'])\n",
    "test1_y = np.array(test1['22']).reshape(1,int(1117500/STEP))[0]\n",
    "test2_pred = np.array(test2['22'])\n",
    "test2_y = np.array(test2['23']).reshape(1,int(1117500/STEP))[0]\n",
    "\n",
    "dev_mse = sum((dev_pred-dev_y)**2)/len(dev_pred)\n",
    "dev_mae = sum(abs(dev_pred-dev_y))/len(dev_pred)\n",
    "test1_mse = sum((test1_pred-test1_y)**2)/len(test1_pred)\n",
    "test1_mae = sum(abs(test1_pred-test1_y))/len(test1_pred)\n",
    "test2_mse = sum((test2_pred-test2_y)**2)/len(test2_pred)\n",
    "test2_mae = sum(abs(test2_pred-test2_y))/len(test2_pred)\n",
    "\n",
    "hist_matrix = [[dev_mse, dev_mae], [test1_mse, test1_mae], [test2_mse, test2_mae]]\n",
    "for i in hist_matrix :\n",
    "    print(str(i[0]) + '/' + str(i[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Constant Correlation Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.295445251397/0.442321381012\n",
      "0.263904487183/0.443618996923\n",
      "0.290317094717/0.457613454021\n"
     ]
    }
   ],
   "source": [
    "pred = sum(dev['20'])/int(1117500/STEP)\n",
    "dev_pred = np.array([pred] * int(1117500/STEP))\n",
    "pred = sum(test1['21'])/int(1117500/STEP)\n",
    "test1_pred = np.array([pred] * int(1117500/STEP))\n",
    "pred = sum(test2['22'])/int(1117500/STEP)\n",
    "test2_pred = np.array([pred] * int(1117500/STEP))\n",
    "\n",
    "dev_mse = sum((dev_pred-dev_y)**2)/len(dev_pred)\n",
    "dev_mae = sum(abs(dev_pred-dev_y))/len(dev_pred)\n",
    "test1_mse = sum((test1_pred-test1_y)**2)/len(test1_pred)\n",
    "test1_mae = sum(abs(test1_pred-test1_y))/len(test1_pred)\n",
    "test2_mse = sum((test2_pred-test2_y)**2)/len(test2_pred)\n",
    "test2_mae = sum(abs(test2_pred-test2_y))/len(test2_pred)\n",
    "\n",
    "cc_matrix = [[dev_mse, dev_mae], [test1_mse, test1_mae], [test2_mse, test2_mae]]\n",
    "for i in cc_matrix :\n",
    "    print(str(i[0]) + '/' + str(i[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Multi Group Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data_df = pd.read_csv('C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/dataset.csv')\n",
    "data_df = data_df.loc[:, ~data_df.columns.str.contains('^Unnamed')]\n",
    "num_list = []\n",
    "for i in range(24):\n",
    "    num_list.append(str(i))\n",
    "data_df = data_df[num_list].copy()\n",
    "data_df = np.transpose(data_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Industrials' 'Health Care' 'Information Technology'\n",
      " 'Consumer Discretionary' 'Utilities' 'Financials' 'Materials'\n",
      " 'Real Estate' 'Consumer Staples' 'Energy' 'Telecommunication Services']\n",
      "   Unnamed: 0             GICS_sector               GICS_sub_industry  \\\n",
      "0           0             Industrials        Industrial Conglomerates   \n",
      "1           1             Health Care           Health Care Equipment   \n",
      "2           2             Health Care                 Pharmaceuticals   \n",
      "3           3  Information Technology  IT Consulting & Other Services   \n",
      "4           4  Information Technology     Home Entertainment Software   \n",
      "\n",
      "               company ticker  \n",
      "0           3M Company    MMM  \n",
      "1  Abbott Laboratories    ABT  \n",
      "2          AbbVie Inc.   ABBV  \n",
      "3        Accenture plc    ACN  \n",
      "4  Activision Blizzard   ATVI  \n"
     ]
    }
   ],
   "source": [
    "data = pd.read_csv('C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/SP500_list.csv')\n",
    "print(data['GICS_sector'].unique())\n",
    "print(data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# portfolio list\n",
    "portfolio = ['CELG', 'PXD', 'WAT', 'LH', 'AMGN', 'AOS', 'EFX', 'CRM', 'NEM', 'JNPR', 'LB', 'CTAS', 'MAT', 'MDLZ', 'VLO', 'APH', 'ADM', 'MLM', 'BK', 'NOV', 'BDX', 'RRC', 'IVZ', 'ED', 'SBUX', 'GRMN', 'CI', 'ZION', 'COO', 'TIF', 'RHT', 'FDX', 'LLL', 'GLW', 'GPN', 'IPGP', 'GPC', 'HPQ', 'ADI', 'AMG', 'MTB', 'YUM', 'SYK', 'KMX', 'AME', 'AAP', 'DAL', 'A', 'MON', 'BRK', 'BMY', 'KMB', 'JPM', 'CCI', 'AET', 'DLTR', 'MGM', 'FL', 'HD', 'CLX', 'OKE', 'UPS', 'WMB', 'IFF', 'CMS', 'ARNC', 'VIAB', 'MMC', 'REG', 'ES', 'ITW', 'NDAQ', 'AIZ', 'VRTX', 'CTL', 'QCOM', 'MSI', 'NKTR', 'AMAT', 'BWA', 'ESRX', 'TXT', 'EXR', 'VNO', 'BBT', 'WDC', 'UAL', 'PVH', 'NOC', 'PCAR', 'NSC', 'UAA', 'FFIV', 'PHM', 'LUV', 'HUM', 'SPG', 'SJM', 'ABT', 'CMG', 'ALK', 'ULTA', 'TMK', 'TAP', 'SCG', 'CAT', 'TMO', 'AES', 'MRK', 'RMD', 'MKC', 'WU', 'ACN', 'HIG', 'TEL', 'DE', 'ATVI', 'O', 'UNM', 'VMC', 'ETFC', 'CMA', 'NRG', 'RHI', 'RE', 'FMC', 'MU', 'CB', 'LNT', 'GE', 'CBS', 'ALGN', 'SNA', 'LLY', 'LEN', 'MAA', 'OMC', 'F', 'APA', 'CDNS', 'SLG', 'HP', 'XLNX', 'SHW', 'AFL', 'STT', 'PAYX', 'AIG', 'FOX', 'MA']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Industrials': ['AOS', 'EFX', 'CTAS', 'FDX', 'LLL', 'AME', 'DAL', 'UPS', 'ARNC', 'ITW', 'TXT', 'UAL', 'NOC', 'PCAR', 'NSC', 'LUV', 'ALK', 'CAT', 'DE', 'RHI', 'GE'], 'Health Care': ['CELG', 'WAT', 'LH', 'AMGN', 'BDX', 'CI', 'COO', 'SYK', 'A', 'BMY', 'AET', 'VRTX', 'NKTR', 'ESRX', 'HUM', 'ABT', 'TMO', 'MRK', 'RMD', 'ALGN', 'LLY'], 'Information Technology': ['CRM', 'JNPR', 'APH', 'RHT', 'GLW', 'GPN', 'IPGP', 'HPQ', 'ADI', 'QCOM', 'MSI', 'AMAT', 'WDC', 'FFIV', 'WU', 'ACN', 'TEL', 'ATVI', 'MU', 'CDNS', 'XLNX', 'PAYX', 'MA'], 'Consumer Discretionary': ['LB', 'MAT', 'SBUX', 'GRMN', 'TIF', 'GPC', 'YUM', 'KMX', 'AAP', 'DLTR', 'MGM', 'FL', 'HD', 'VIAB', 'BWA', 'PVH', 'UAA', 'PHM', 'CMG', 'ULTA', 'CBS', 'SNA', 'LEN', 'OMC', 'F', 'FOX'], 'Utilities': ['ED', 'CMS', 'ES', 'SCG', 'AES', 'NRG', 'LNT'], 'Financials': ['BK', 'IVZ', 'ZION', 'AMG', 'MTB', 'BRK', 'JPM', 'MMC', 'NDAQ', 'AIZ', 'BBT', 'TMK', 'HIG', 'UNM', 'ETFC', 'CMA', 'RE', 'CB', 'AFL', 'STT', 'AIG'], 'Materials': ['NEM', 'MLM', 'MON', 'IFF', 'VMC', 'FMC', 'SHW'], 'Real Estate': ['CCI', 'REG', 'EXR', 'VNO', 'SPG', 'O', 'MAA', 'SLG'], 'Consumer Staples': ['MDLZ', 'ADM', 'KMB', 'CLX', 'SJM', 'TAP', 'MKC'], 'Energy': ['PXD', 'VLO', 'NOV', 'RRC', 'OKE', 'WMB', 'APA', 'HP'], 'Telecommunication Services': ['CTL']}\n"
     ]
    }
   ],
   "source": [
    "pf_sector_item = {'Industrials':[],\n",
    "                  'Health Care':[],\n",
    "                  'Information Technology':[],\n",
    "                  'Consumer Discretionary':[],\n",
    "                  'Utilities':[],\n",
    "                  'Financials' :[],\n",
    "                  'Materials':[],\n",
    "                  'Real Estate':[],\n",
    "                  'Consumer Staples':[],\n",
    "                  'Energy':[],\n",
    "                  'Telecommunication Services':[]}\n",
    "for item in portfolio :\n",
    "    pf_sector_item[data[data.ticker == item]['GICS_sector'].values[0]] = pf_sector_item[data[data.ticker == item]['GICS_sector'].values[0]]+[item]\n",
    "print(pf_sector_item)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "market_data = pd.read_csv('C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/stock08_price.csv')\n",
    "pf_sector_dev = {}\n",
    "pf_sector_test1 = {}\n",
    "pf_sector_test2 = {}\n",
    "\n",
    "for i in range(150):\n",
    "    for j in range(149-i):\n",
    "        a = portfolio[i]\n",
    "        b = portfolio[149-j]\n",
    "        a_price = market_data[a]\n",
    "        b_price = market_data[b]\n",
    "        a_sector = data[data.ticker == a]['GICS_sector'].values[0]\n",
    "        b_sector = data[data.ticker == b]['GICS_sector'].values[0]\n",
    "        sector_pair = max(a_sector, b_sector)+'_'+min(a_sector, b_sector)\n",
    "        \n",
    "        dev = []\n",
    "        test1 = []\n",
    "        test2 = []\n",
    "        for k in range(5):\n",
    "            dev_start = 2000 + k*20\n",
    "            test1_start = 2100 + k*20\n",
    "            test2_start = 2200 + k*20\n",
    "            dev.append(a_price[dev_start:dev_start+100].corr(b_price[dev_start:dev_start+100]))\n",
    "            test1.append(a_price[test1_start:test1_start+100].corr(b_price[test1_start:test1_start+100]))\n",
    "            test2.append(a_price[test2_start:test2_start+100].corr(b_price[test2_start:test2_start+100]))\n",
    "        \n",
    "        try:\n",
    "            pf_sector_dev[sector_pair] = pf_sector_dev[sector_pair] + [dev]\n",
    "        except KeyError :\n",
    "            pf_sector_dev[sector_pair] = [dev]\n",
    "            \n",
    "        try:\n",
    "            pf_sector_test1[sector_pair] = pf_sector_test1[sector_pair] + [test1]\n",
    "        except KeyError :\n",
    "            pf_sector_test1[sector_pair] = [test1]\n",
    "            \n",
    "        try:\n",
    "            pf_sector_test2[sector_pair] = pf_sector_test2[sector_pair] + [test2]\n",
    "        except KeyError :\n",
    "            pf_sector_test2[sector_pair] = [test2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pairs = [key for key in pf_sector_dev]\n",
    "sector_pair_corr_dev = {}\n",
    "sector_pair_corr_test1 = {}\n",
    "sector_pair_corr_test2 = {}\n",
    "for pair in pairs :\n",
    "    dev_zeroes = np.array([0] * 5)\n",
    "    test1_zeroes = np.array([0] * 5)\n",
    "    test2_zeroes = np.array([0] * 5)\n",
    "    dev_length = len(pf_sector_dev[pair])\n",
    "    test1_length = len(pf_sector_test1[pair])\n",
    "    test2_length = len(pf_sector_test2[pair])\n",
    "    for arr in pf_sector_dev[pair] :\n",
    "        dev_zeroes = dev_zeroes + np.array(arr)\n",
    "        dev_result = dev_zeroes/dev_length\n",
    "    for arr in pf_sector_test1[pair] :\n",
    "        test1_zeroes = test1_zeroes + np.array(arr)\n",
    "        test1_result = test1_zeroes/test1_length\n",
    "    for arr in pf_sector_test2[pair] :\n",
    "        test2_zeroes = test2_zeroes + np.array(arr)\n",
    "        test2_result = test2_zeroes/test2_length\n",
    "    sector_pair_corr_dev[pair] = dev_result\n",
    "    sector_pair_corr_test1[pair] = test1_result\n",
    "    sector_pair_corr_test2[pair] = test2_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.307930389914/0.451498858397\n",
      "0.291014800661/0.455505188547\n",
      "0.287446079292/0.447963398344\n"
     ]
    }
   ],
   "source": [
    "num_list = [STEP*i for i in range(int(1117500/STEP))]\n",
    "dataset = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/dataset.csv\")\n",
    "dev_y = dataset['21'].copy()\n",
    "test1_y = dataset['22'].copy()\n",
    "test2_y = dataset['23'].copy()\n",
    "\n",
    "dev_y = np.array(dev_y[num_list]).reshape(1,int(1117500/STEP))[0]\n",
    "test1_y = np.array(test1_y[num_list]).reshape(1,int(1117500/STEP))[0]\n",
    "test2_y = np.array(test2_y[num_list]).reshape(1,int(1117500/STEP))[0]\n",
    "\n",
    "dev_pred = []\n",
    "test1_pred = []\n",
    "test2_pred = []\n",
    "for i in range(150):\n",
    "    for j in range(149-i):\n",
    "        a = portfolio[i]\n",
    "        b = portfolio[149-j]\n",
    "        a_sector = data[data.ticker == a]['GICS_sector'].values[0]\n",
    "        b_sector = data[data.ticker == b]['GICS_sector'].values[0]\n",
    "        sector_pair = max(a_sector, b_sector)+'_'+min(a_sector, b_sector)\n",
    "        \n",
    "        dev_pred = dev_pred + list(sector_pair_corr_dev[sector_pair])\n",
    "        test1_pred = test1_pred + list(sector_pair_corr_test1[sector_pair])\n",
    "        test2_pred = test2_pred + list(sector_pair_corr_test2[sector_pair])\n",
    "dev_pred = np.array(dev_pred)\n",
    "test1_pred = np.array(test1_pred)\n",
    "test2_pred = np.array(test2_pred)\n",
    "\n",
    "\n",
    "dev_mse = sum((dev_pred-dev_y)**2)/len(dev_pred)\n",
    "dev_mae = sum(abs(dev_pred-dev_y))/len(dev_pred)\n",
    "test1_mse = sum((test1_pred-test1_y)**2)/len(test1_pred)\n",
    "test1_mae = sum(abs(test1_pred-test1_y))/len(test1_pred)\n",
    "test2_mse = sum((test2_pred-test2_y)**2)/len(test2_pred)\n",
    "test2_mae = sum(abs(test2_pred-test2_y))/len(test2_pred)\n",
    "\n",
    "mg_matrix = [[dev_mse, dev_mae], [test1_mse, test1_mae], [test2_mse, test2_mae]]\n",
    "for i in mg_matrix :\n",
    "    print(str(i[0]) + '/' + str(i[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Single Index Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data_df = pd.read_csv('C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/stock08_price.csv')\n",
    "data_df = data_df.loc[:, ~data_df.columns.str.contains('^Unnamed')]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.403468057827/0.51653447995\n",
      "0.351679806968/0.492034991305\n",
      "0.385985603724/0.50089471591\n"
     ]
    }
   ],
   "source": [
    "dev_pred = []\n",
    "test1_pred = []\n",
    "test2_pred = []\n",
    "\n",
    "for i in range(150):\n",
    "    for j in range(149-i):\n",
    "        a = portfolio[i]\n",
    "        b = portfolio[149-j]\n",
    "        for k in range(5):\n",
    "            dev_start = 2000 + k*20\n",
    "            test1_start = 2100 + k*20\n",
    "            test2_start = 2200 + k*20\n",
    "            dev_pred.append(data_df[a][dev_start:dev_start+100].corr(data_df['SP500'][dev_start:dev_start+100]) *\n",
    "                            data_df[b][dev_start:dev_start+100].corr(data_df['SP500'][dev_start:dev_start+100]))\n",
    "            test1_pred.append(data_df[a][test1_start:test1_start+100].corr(data_df['SP500'][test1_start:test1_start+100])*\n",
    "                              data_df[b][test1_start:test1_start+100].corr(data_df['SP500'][test1_start:test1_start+100]))\n",
    "            test2_pred.append(data_df[a][test2_start:test2_start+100].corr(data_df['SP500'][test2_start:test2_start+100])*\n",
    "                              data_df[b][test2_start:test2_start+100].corr(data_df['SP500'][test2_start:test2_start+100]))\n",
    "dev_pred = np.array(dev_pred)\n",
    "test1_pred = np.array(test1_pred)\n",
    "test2_pred = np.array(test2_pred)\n",
    "            \n",
    "num_list = [STEP*i for i in range(int(1117500/STEP))]\n",
    "dataset = pd.read_csv(\"C:/Users/Froilan/Desktop/myFiles/JupyterFiles/stock_correlation_prediction/dataset.csv\")\n",
    "dev_y = dataset['21'].copy()\n",
    "test1_y = dataset['22'].copy()\n",
    "test2_y = dataset['23'].copy()\n",
    "\n",
    "dev_y = np.array(dev_y[num_list]).reshape(1,int(1117500/STEP))[0]\n",
    "test1_y = np.array(test1_y[num_list]).reshape(1,int(1117500/STEP))[0]\n",
    "test2_y = np.array(test2_y[num_list]).reshape(1,int(1117500/STEP))[0]\n",
    "\n",
    "  \n",
    "\n",
    "dev_mse = sum((dev_pred-dev_y)**2)/len(dev_pred)\n",
    "dev_mae = sum(abs(dev_pred-dev_y))/len(dev_pred)\n",
    "test1_mse = sum((test1_pred-test1_y)**2)/len(test1_pred)\n",
    "test1_mae = sum(abs(test1_pred-test1_y))/len(test1_pred)\n",
    "test2_mse = sum((test2_pred-test2_y)**2)/len(test2_pred)\n",
    "test2_mae = sum(abs(test2_pred-test2_y))/len(test2_pred)\n",
    "\n",
    "mg_matrix = [[dev_mse, dev_mae], [test1_mse, test1_mae], [test2_mse, test2_mae]]\n",
    "for i in mg_matrix :\n",
    "    print(str(i[0]) + '/' + str(i[1]))                  "
   ]
  },
  {
   "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": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "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": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "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": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}