4_3_1__Stock_Picking_Predicting_The_Future_Forecasting_ARIMA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# STOCK PICKING 3.1 - FORECASTING ARIMAs \n",
    "\n",
    "_In mental hospitals there are two kinds of people: who they think to be Napoleon and who they think they can estimate confidence in predicting stock prices_ (Gino Tesei). "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Getting data\n",
    "\n",
    "Let's consider all the large caps listed on NASDAQ and NYSE (~1000 stocks). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEICAYAAABS0fM3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDLOc/w/E4D3FNmpUTwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1ff4773f3c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt\n",
    "import pandas_datareader.data as web\n",
    "import datetime\n",
    "\n",
    "def get_data(symbols, \n",
    "             add_ref=True,\n",
    "             data_source='yahoo',\n",
    "             price='Adj Close',\n",
    "             start='1/21/2010', \n",
    "             end='4/15/2016'):\n",
    "    \"\"\"Read stock data (adjusted close) for given symbols from.\"\"\"\n",
    "    \n",
    "    if add_ref and 'SPY' not in symbols:  # add SPY for reference, if absent\n",
    "        symbols.insert(0, 'SPY')\n",
    "\n",
    "    df = web.DataReader(symbols, \n",
    "                        data_source=data_source,\n",
    "                        start=start, \n",
    "                        end=end)\n",
    "    \n",
    "    return df[price,:,:]\n",
    "\n",
    "def compute_daily_returns(df):\n",
    "    \"\"\"Compute and return the daily return values.\"\"\"\n",
    "    # Note: Returned DataFrame must have the same number of rows\n",
    "    daily_returns = (df / df.shift(1)) - 1 \n",
    "    daily_returns.ix[0,:] = 0 \n",
    "    return daily_returns\n",
    "\n",
    "def fill_missing_values(df_data):\n",
    "    \"\"\"Fill missing values in data frame, in place.\"\"\"\n",
    "    df_data.fillna(method='ffill',inplace=True)\n",
    "    df_data.fillna(method='backfill',inplace=True)\n",
    "    return df_data\n",
    "\n",
    "def cumulative_returns(df):\n",
    "    return df/df.ix[0,:] - 1 \n",
    "\n",
    "df = fill_missing_values(get_data(symbols=['GOOG','SPY','IBM','GLD'],\n",
    "                             start='4/21/2015', \n",
    "                             end='7/15/2016'))\n",
    "df.plot()\n",
    "plt.show(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>GLD</th>\n",
       "      <th>GOOG</th>\n",
       "      <th>IBM</th>\n",
       "      <th>SPY</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-04-21</th>\n",
       "      <td>115.379997</td>\n",
       "      <td>533.972413</td>\n",
       "      <td>151.610189</td>\n",
       "      <td>201.067670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-04-22</th>\n",
       "      <td>113.830002</td>\n",
       "      <td>539.367458</td>\n",
       "      <td>152.625483</td>\n",
       "      <td>202.055740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-04-23</th>\n",
       "      <td>114.660004</td>\n",
       "      <td>547.002472</td>\n",
       "      <td>157.129675</td>\n",
       "      <td>202.564164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-04-24</th>\n",
       "      <td>113.050003</td>\n",
       "      <td>565.062561</td>\n",
       "      <td>156.705093</td>\n",
       "      <td>203.034208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-04-27</th>\n",
       "      <td>115.330002</td>\n",
       "      <td>555.369995</td>\n",
       "      <td>157.581930</td>\n",
       "      <td>202.190041</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   GLD        GOOG         IBM         SPY\n",
       "Date                                                      \n",
       "2015-04-21  115.379997  533.972413  151.610189  201.067670\n",
       "2015-04-22  113.830002  539.367458  152.625483  202.055740\n",
       "2015-04-23  114.660004  547.002472  157.129675  202.564164\n",
       "2015-04-24  113.050003  565.062561  156.705093  203.034208\n",
       "2015-04-27  115.330002  555.369995  157.581930  202.190041"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "simbols = [\"A\",\"AA\",\"AAL\",\"AAP\",\"AAPL\",\"ABB\",\"ABBV\",\"ABC\",\"ABMD\",\"ABT\",\"ABX\",\"ACC\",\"ACGL\",\"ACM\",\"ACN\",\"ADBE\",\"ADI\",\"ADM\",\"ADNT\",\"ADP\",\"ADS\",\"ADSK\",\"AEE\",\"AEG\",\"AEM\",\"AEP\",\"AER\",\"AES\",\"AET\",\"AFG\",\"AFL\",\"AGCO\",\"AGN\",\"AGNC\",\"AGR\",\"AGU\",\"AIG\",\"AIV\",\"AIZ\",\"AJG\",\"AKAM\",\"ALB\",\"ALGN\",\"ALK\",\"ALKS\",\"ALL\",\"ALLE\",\"ALLY\",\"ALSN\",\"ALV\",\"ALXN\",\"AM\",\"AMAT\",\"AMD\",\"AME\",\"AMG\",\"AMGN\",\"AMH\",\"AMP\",\"AMT\",\"AMTD\",\"AMX\",\"AMZN\",\"ANET\",\"ANSS\",\"ANTM\",\"AON\",\"AOS\",\"APA\",\"APC\",\"APD\",\"APH\",\"APO\",\"AR\",\"ARCC\",\"ARD\",\"ARE\",\"ARMK\",\"ARNC\",\"ARRS\",\"ARW\",\"ASH\",\"ASML\",\"ASR\",\"ASX\",\"ATH\",\"ATO\",\"ATR\",\"ATVI\",\"AVB\",\"AVGO\",\"AVY\",\"AWK\",\"AXP\",\"AXS\",\"AXTA\",\"AYI\",\"AZN\",\"AZO\",\"BA\",\"BABA\",\"BAC\",\"BAH\",\"BAM\",\"BAP\",\"BAX\",\"BBBY\",\"BBD\",\"BBL\",\"BBRY\",\"BBT\",\"BBVA\",\"BBY\",\"BC\",\"BCE\",\"BCH\",\"BCR\",\"BCS\",\"BDX\",\"BEN\",\"BERY\",\"BF.B\",\"BG\",\"BHI\",\"BHP\",\"BIDU\",\"BIIB\",\"BIO\",\"BIP\",\"BIVV\",\"BK\",\"BKFS\",\"BLK\",\"BLL\",\"BMA\",\"BMO\",\"BMRN\",\"BMY\",\"BNS\",\"BOKF\",\"BP\",\"BPL\",\"BPY\",\"BR\",\"BRCD\",\"BRFS\",\"BRK.A\",\"BRO\",\"BRX\",\"BSAC\",\"BSBR\",\"BSMX\",\"BSX\",\"BT\",\"BUD\",\"BURL\",\"BWA\",\"BX\",\"BXP\",\"C\",\"CA\",\"CAG\",\"CAH\",\"CAJ\",\"CAT\",\"CB\",\"CBG\",\"CBOE\",\"CBS\",\"CBSH\",\"CC\",\"CCE\",\"CCI\",\"CCK\",\"CCL\",\"CDK\",\"CDNS\",\"CDW\",\"CE\",\"CELG\",\"CEO\",\"CERN\",\"CF\",\"CFG\",\"CFR\",\"CG\",\"CGNX\",\"CHA\",\"CHD\",\"CHK\",\"CHKP\",\"CHL\",\"CHRW\",\"CHT\",\"CHTR\",\"CHU\",\"CI\",\"CINF\",\"CIT\",\"CL\",\"CLNS\",\"CLR\",\"CLX\",\"CM\",\"CMA\",\"CMCSA\",\"CME\",\"CMG\",\"CMI\",\"CMS\",\"CNA\",\"CNC\",\"CNHI\",\"CNI\",\"CNK\",\"CNP\",\"CNQ\",\"COF\",\"COG\",\"COH\",\"COHR\",\"COL\",\"COMM\",\"COO\",\"COP\",\"COST\",\"COTY\",\"CP\",\"CPA\",\"CPB\",\"CPL\",\"CPRT\",\"CPT\",\"CRH\",\"CRM\",\"CS\",\"CSCO\",\"CSGP\",\"CSL\",\"CSX\",\"CTAS\",\"CTL\",\"CTRP\",\"CTSH\",\"CTXS\",\"CUK\",\"CVE\",\"CVS\",\"CVX\",\"CX\",\"CXO\",\"D\",\"DAL\",\"DB\",\"DCI\",\"DCM\",\"DCP\",\"DD\",\"DE\",\"DEI\",\"DEO\",\"DFS\",\"DG\",\"DGX\",\"DHI\",\"DHR\",\"DIS\",\"DISH\",\"DKS\",\"DLB\",\"DLPH\",\"DLR\",\"DLTR\",\"DNKN\",\"DOV\",\"DOW\",\"DOX\",\"DPS\",\"DPZ\",\"DRE\",\"DRI\",\"DTE\",\"DUK\",\"DVA\",\"DVMT\",\"DVN\",\"DXC\",\"DXCM\",\"E\",\"EA\",\"EBAY\",\"EC\",\"ECA\",\"ECL\",\"ED\",\"EDU\",\"EEP\",\"EFX\",\"EGN\",\"EIX\",\"EL\",\"ELS\",\"EMN\",\"EMR\",\"ENB\",\"ENBL\",\"ENIA\",\"ENIC\",\"ENLK\",\"EOCC\",\"EOG\",\"EPD\",\"EPR\",\"EQGP\",\"EQIX\",\"EQM\",\"EQR\",\"EQT\",\"ERIC\",\"ERIE\",\"ES\",\"ESLT\",\"ESRX\",\"ESS\",\"ETE\",\"ETFC\",\"ETN\",\"ETR\",\"EV\",\"EVHC\",\"EW\",\"EWBC\",\"EXC\",\"EXEL\",\"EXPD\",\"EXPE\",\"EXR\",\"F\",\"FANG\",\"FAST\",\"FB\",\"FBHS\",\"FBR\",\"FCE.A\",\"FCX\",\"FDC\",\"FDS\",\"FDX\",\"FE\",\"FFIV\",\"FIS\",\"FISV\",\"FITB\",\"FL\",\"FLEX\",\"FLIR\",\"FLR\",\"FLS\",\"FLT\",\"FMC\",\"FMS\",\"FMX\",\"FNF\",\"FOXA\",\"FRC\",\"FRT\",\"FTI\",\"FTNT\",\"FTV\",\"G\",\"GD\",\"GDDY\",\"GE\",\"GG\",\"GGG\",\"GGP\",\"GIB\",\"GIL\",\"GILD\",\"GIS\",\"GLPI\",\"GLW\",\"GM\",\"GNTX\",\"GOLD\",\"GOOGL\",\"GPC\",\"GPN\",\"GPS\",\"GRMN\",\"GS\",\"GSK\",\"GT\",\"GWW\",\"GXP\",\"H\",\"HAL\",\"HAS\",\"HBAN\",\"HBI\",\"HCA\",\"HCN\",\"HCP\",\"HD\",\"HDB\",\"HDS\",\"HES\",\"HFC\",\"HHC\",\"HIG\",\"HII\",\"HIW\",\"HLF\",\"HLT\",\"HMC\",\"HOG\",\"HOLX\",\"HON\",\"HP\",\"HPE\",\"HPP\",\"HPQ\",\"HRB\",\"HRL\",\"HRS\",\"HSBC\",\"HSIC\",\"HST\",\"HSY\",\"HTA\",\"HTHT\",\"HUBB\",\"HUM\",\"HUN\",\"IAC\",\"IBKR\",\"IBM\",\"IBN\",\"ICE\",\"IDXX\",\"IEP\",\"IEX\",\"IFF\",\"IHG\",\"ILMN\",\"INCY\",\"INFO\",\"INFY\",\"ING\",\"INGR\",\"INTC\",\"INTU\",\"INVH\",\"IONS\",\"IP\",\"IPG\",\"IPGP\",\"IR\",\"IRM\",\"ISRG\",\"IT\",\"ITUB\",\"ITW\",\"IVZ\",\"IX\",\"JAZZ\",\"JBHT\",\"JBL\",\"JBLU\",\"JD\",\"JEC\",\"JHX\",\"JKHY\",\"JLL\",\"JNJ\",\"JNPR\",\"JPM\",\"JWN\",\"K\",\"KAR\",\"KB\",\"KEP\",\"KEY\",\"KEYS\",\"KHC\",\"KIM\",\"KKR\",\"KLAC\",\"KMB\",\"KMI\",\"KMX\",\"KO\",\"KORS\",\"KR\",\"KRC\",\"KSS\",\"KSU\",\"KT\",\"KYO\",\"L\",\"LAMR\",\"LAZ\",\"LB\",\"LBTYA\",\"LDOS\",\"LEA\",\"LECO\",\"LEG\",\"LEN\",\"LFC\",\"LFL\",\"LH\",\"LII\",\"LKQ\",\"LLL\",\"LLY\",\"LMT\",\"LN\",\"LNC\",\"LNT\",\"LOGI\",\"LOGM\",\"LOW\",\"LPL\",\"LPT\",\"LRCX\",\"LUK\",\"LULU\",\"LUV\",\"LUX\",\"LVLT\",\"LVS\",\"LW\",\"LYB\",\"LYG\",\"LYV\",\"M\",\"MA\",\"MAA\",\"MAC\",\"MAN\",\"MAR\",\"MAS\",\"MAT\",\"MBLY\",\"MBT\",\"MCD\",\"MCHP\",\"MCK\",\"MCO\",\"MD\",\"MDLZ\",\"MDT\",\"MDU\",\"MELI\",\"MET\",\"MFC\",\"MFG\",\"MGA\",\"MGM\",\"MHK\",\"MIC\",\"MIDD\",\"MJN\",\"MKC\",\"MKL\",\"MKTX\",\"MLCO\",\"MLM\",\"MMC\",\"MMM\",\"MMP\",\"MNST\",\"MO\",\"MOMO\",\"MON\",\"MOS\",\"MPC\",\"MPLX\",\"MRK\",\"MRO\",\"MRVL\",\"MS\",\"MSCC\",\"MSCI\",\"MSFT\",\"MSI\",\"MSM\",\"MT\",\"MTB\",\"MTD\",\"MTN\",\"MTU\",\"MU\",\"MXIM\",\"MYL\",\"NBL\",\"NCLH\",\"NCR\",\"NDAQ\",\"NDSN\",\"NEE\",\"NEM\",\"NEU\",\"NFLX\",\"NFX\",\"NGG\",\"NI\",\"NKE\",\"NLSN\",\"NLY\",\"NMR\",\"NNN\",\"NOC\",\"NOK\",\"NOV\",\"NOW\",\"NRZ\",\"NSC\",\"NTAP\",\"NTES\",\"NTRS\",\"NUAN\",\"NUE\",\"NVDA\",\"NVO\",\"NVR\",\"NVS\",\"NWL\",\"NXPI\",\"NYCB\",\"O\",\"OA\",\"OAK\",\"OC\",\"ODFL\",\"OGE\",\"OHI\",\"OKE\",\"OKS\",\"OLED\",\"OLN\",\"OMC\",\"ON\",\"ORAN\",\"ORCL\",\"ORI\",\"ORLY\",\"OSK\",\"OTEX\",\"OXY\",\"OZRK\",\"PAA\",\"PAC\",\"PACW\",\"PAGP\",\"PANW\",\"PAYX\",\"PBCT\",\"PBR\",\"PCAR\",\"PCG\",\"PCLN\",\"PE\",\"PEG\",\"PEP\",\"PF\",\"PFE\",\"PFG\",\"PG\",\"PGR\",\"PH\",\"PHG\",\"PHI\",\"PHM\",\"PII\",\"PK\",\"PKG\",\"PKI\",\"PKX\",\"PLD\",\"PM\",\"PNC\",\"PNR\",\"PNRA\",\"PNW\",\"POOL\",\"POST\",\"POT\",\"PPC\",\"PPG\",\"PPL\",\"PRGO\",\"PRU\",\"PSA\",\"PSO\",\"PSX\",\"PSXP\",\"PTC\",\"PTR\",\"PUK\",\"PVH\",\"PWR\",\"PX\",\"PXD\",\"PYPL\",\"Q\",\"QCOM\",\"QGEN\",\"QRVO\",\"QVCA\",\"RACE\",\"RAI\",\"RBS\",\"RCI\",\"RCL\",\"RDS.A\",\"RDY\",\"RE\",\"REG\",\"REGN\",\"RELX\",\"RENX\",\"RF\",\"RGA\",\"RHI\",\"RHT\",\"RIO\",\"RJF\",\"RL\",\"RMD\",\"RNR\",\"ROK\",\"ROL\",\"ROP\",\"ROST\",\"RPM\",\"RRC\",\"RS\",\"RSG\",\"RSPP\",\"RTN\",\"RY\",\"RYAAY\",\"S\",\"SABR\",\"SAN\",\"SAP\",\"SATS\",\"SBAC\",\"SBNY\",\"SBS\",\"SBUX\",\"SCCO\",\"SCG\",\"SCHW\",\"SCI\",\"SEE\",\"SEIC\",\"SEP\",\"SERV\",\"SGEN\",\"SHG\",\"SHLX\",\"SHOP\",\"SHPG\",\"SHW\",\"SINA\",\"SIRI\",\"SIVB\",\"SIX\",\"SJM\",\"SJR\",\"SKM\",\"SLB\",\"SLF\",\"SLG\",\"SLM\",\"SLW\",\"SMFG\",\"SMG\",\"SMI\",\"SNA\",\"SNAP\",\"SNE\",\"SNI\",\"SNN\",\"SNP\",\"SNPS\",\"SNV\",\"SNY\",\"SO\",\"SON\",\"SPB\",\"SPG\",\"SPGI\",\"SPLK\",\"SPLS\",\"SPR\",\"SQ\",\"SRCL\",\"SRE\",\"SSL\",\"SSNC\",\"ST\",\"STE\",\"STI\",\"STLD\",\"STM\",\"STO\",\"STT\",\"STWD\",\"STX\",\"STZ\",\"SU\",\"SUI\",\"SWK\",\"SWKS\",\"SYF\",\"SYK\",\"SYMC\",\"SYT\",\"SYY\",\"T\",\"TAL\",\"TAP\",\"TD\",\"TDG\",\"TEAM\",\"TECK\",\"TEF\",\"TEL\",\"TER\",\"TEVA\",\"TFX\",\"TGNA\",\"TGT\",\"TI\",\"TIF\",\"TJX\",\"TKC\",\"TLK\",\"TLLP\",\"TM\",\"TMK\",\"TMO\",\"TMUS\",\"TOL\",\"TOT\",\"TRGP\",\"TRI\",\"TRIP\",\"TRMB\",\"TROW\",\"TRP\",\"TRQ\",\"TRU\",\"TRV\",\"TS\",\"TSCO\",\"TSLA\",\"TSM\",\"TSN\",\"TSO\",\"TSRO\",\"TSS\",\"TSU\",\"TTC\",\"TTM\",\"TTWO\",\"TU\",\"TV\",\"TWTR\",\"TWX\",\"TXN\",\"TXT\",\"TYL\",\"UAL\",\"UBS\",\"UDR\",\"UGI\",\"UGP\",\"UHAL\",\"UHS\",\"UL\",\"ULTA\",\"ULTI\",\"UMC\",\"UN\",\"UNH\",\"UNM\",\"UNP\",\"UPS\",\"URI\",\"USB\",\"USFD\",\"UTHR\",\"UTX\",\"V\",\"VAL\",\"VALE\",\"VAR\",\"VEDL\",\"VEEV\",\"VEON\",\"VER\",\"VFC\",\"VIAB\",\"VIPS\",\"VIV\",\"VLO\",\"VMC\",\"VMW\",\"VNO\",\"VNTV\",\"VOD\",\"VOYA\",\"VRSK\",\"VRSN\",\"VRTX\",\"VTR\",\"VZ\",\"W\",\"WAB\",\"WAL\",\"WAT\",\"WB\",\"WBA\",\"WBC\",\"WBK\",\"WCG\",\"WCN\",\"WDAY\",\"WDC\",\"WEC\",\"WES\",\"WF\",\"WFC\",\"WFM\",\"WFT\",\"WGP\",\"WHR\",\"WIT\",\"WLK\",\"WLTW\",\"WM\",\"WMB\",\"WMT\",\"WOOF\",\"WPC\",\"WPPGY\",\"WPZ\",\"WR\",\"WRB\",\"WRK\",\"WST\",\"WTR\",\"WU\",\"WUBA\",\"WY\",\"WYN\",\"WYNN\",\"XEC\",\"XEL\",\"XL\",\"XLNX\",\"XOM\",\"XPO\",\"XRAY\",\"XRX\",\"XYL\",\"Y\",\"YHOO\",\"YNDX\",\"YPF\",\"YUM\",\"YUMC\",\"ZAYO\",\"ZBH\",\"ZBRA\",\"ZION\",\"ZTO\",\"ZTS\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "983"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(simbols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "start_date 03-28-2017\n"
     ]
    }
   ],
   "source": [
    "DELTA = 45 ## delay in days \n",
    "start_date = (datetime.date.today() - datetime.timedelta(DELTA)).strftime(\"%m-%d-%Y\")\n",
    "print(\"start_date\",start_date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "end_date 05-12-2017\n"
     ]
    }
   ],
   "source": [
    "end_date = (datetime.date.today()).strftime(\"%m-%d-%Y\")\n",
    "print(\"end_date\",end_date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\gtesei\\AppData\\Local\\Continuum\\Miniconda3\\lib\\site-packages\\pandas_datareader\\base.py:192: SymbolWarning: Failed to read symbol: 'BF.B', replacing with NaN.\n",
      "  warnings.warn(msg.format(sym), SymbolWarning)\n",
      "C:\\Users\\gtesei\\AppData\\Local\\Continuum\\Miniconda3\\lib\\site-packages\\pandas_datareader\\base.py:192: SymbolWarning: Failed to read symbol: 'BRK.A', replacing with NaN.\n",
      "  warnings.warn(msg.format(sym), SymbolWarning)\n",
      "C:\\Users\\gtesei\\AppData\\Local\\Continuum\\Miniconda3\\lib\\site-packages\\pandas_datareader\\base.py:192: SymbolWarning: Failed to read symbol: 'FCE.A', replacing with NaN.\n",
      "  warnings.warn(msg.format(sym), SymbolWarning)\n",
      "C:\\Users\\gtesei\\AppData\\Local\\Continuum\\Miniconda3\\lib\\site-packages\\pandas_datareader\\base.py:192: SymbolWarning: Failed to read symbol: 'RDS.A', replacing with NaN.\n",
      "  warnings.warn(msg.format(sym), SymbolWarning)\n"
     ]
    }
   ],
   "source": [
    "df = fill_missing_values(get_data(symbols=simbols,\n",
    "                             start=start_date, \n",
    "                             end=end_date))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(32, 984)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>AA</th>\n",
       "      <th>AAL</th>\n",
       "      <th>AAP</th>\n",
       "      <th>AAPL</th>\n",
       "      <th>ABB</th>\n",
       "      <th>ABBV</th>\n",
       "      <th>ABC</th>\n",
       "      <th>ABMD</th>\n",
       "      <th>ABT</th>\n",
       "      <th>...</th>\n",
       "      <th>YNDX</th>\n",
       "      <th>YPF</th>\n",
       "      <th>YUM</th>\n",
       "      <th>YUMC</th>\n",
       "      <th>ZAYO</th>\n",
       "      <th>ZBH</th>\n",
       "      <th>ZBRA</th>\n",
       "      <th>ZION</th>\n",
       "      <th>ZTO</th>\n",
       "      <th>ZTS</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-03-28</th>\n",
       "      <td>53.307702</td>\n",
       "      <td>32.919998</td>\n",
       "      <td>42.599998</td>\n",
       "      <td>148.360001</td>\n",
       "      <td>143.208894</td>\n",
       "      <td>22.665505</td>\n",
       "      <td>65.330055</td>\n",
       "      <td>87.500000</td>\n",
       "      <td>125.050003</td>\n",
       "      <td>44.428253</td>\n",
       "      <td>...</td>\n",
       "      <td>22.430000</td>\n",
       "      <td>23.900000</td>\n",
       "      <td>64.168410</td>\n",
       "      <td>26.549999</td>\n",
       "      <td>32.900002</td>\n",
       "      <td>120.950002</td>\n",
       "      <td>88.580002</td>\n",
       "      <td>41.299999</td>\n",
       "      <td>12.88</td>\n",
       "      <td>54.053044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-29</th>\n",
       "      <td>52.988494</td>\n",
       "      <td>32.910000</td>\n",
       "      <td>41.959999</td>\n",
       "      <td>149.669998</td>\n",
       "      <td>143.527571</td>\n",
       "      <td>22.588149</td>\n",
       "      <td>65.122117</td>\n",
       "      <td>89.510002</td>\n",
       "      <td>124.879997</td>\n",
       "      <td>44.398432</td>\n",
       "      <td>...</td>\n",
       "      <td>22.240000</td>\n",
       "      <td>23.850000</td>\n",
       "      <td>63.750376</td>\n",
       "      <td>26.610001</td>\n",
       "      <td>32.910000</td>\n",
       "      <td>121.370003</td>\n",
       "      <td>88.830002</td>\n",
       "      <td>41.320000</td>\n",
       "      <td>13.27</td>\n",
       "      <td>53.713715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-30</th>\n",
       "      <td>53.188000</td>\n",
       "      <td>33.490002</td>\n",
       "      <td>42.540001</td>\n",
       "      <td>148.669998</td>\n",
       "      <td>143.338349</td>\n",
       "      <td>22.568809</td>\n",
       "      <td>64.815174</td>\n",
       "      <td>89.139999</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>44.279163</td>\n",
       "      <td>...</td>\n",
       "      <td>22.059999</td>\n",
       "      <td>24.180000</td>\n",
       "      <td>63.720510</td>\n",
       "      <td>26.520000</td>\n",
       "      <td>32.810001</td>\n",
       "      <td>121.360001</td>\n",
       "      <td>89.559998</td>\n",
       "      <td>42.380001</td>\n",
       "      <td>13.01</td>\n",
       "      <td>53.364408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-31</th>\n",
       "      <td>52.869999</td>\n",
       "      <td>34.400002</td>\n",
       "      <td>42.299999</td>\n",
       "      <td>148.259995</td>\n",
       "      <td>143.069470</td>\n",
       "      <td>22.626826</td>\n",
       "      <td>64.518133</td>\n",
       "      <td>88.500000</td>\n",
       "      <td>125.199997</td>\n",
       "      <td>44.140015</td>\n",
       "      <td>...</td>\n",
       "      <td>21.930000</td>\n",
       "      <td>24.280001</td>\n",
       "      <td>63.601077</td>\n",
       "      <td>27.200001</td>\n",
       "      <td>32.900002</td>\n",
       "      <td>122.110001</td>\n",
       "      <td>91.250000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>13.09</td>\n",
       "      <td>53.264603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-04-03</th>\n",
       "      <td>53.200001</td>\n",
       "      <td>34.220001</td>\n",
       "      <td>42.450001</td>\n",
       "      <td>145.500000</td>\n",
       "      <td>143.109299</td>\n",
       "      <td>22.472113</td>\n",
       "      <td>64.389409</td>\n",
       "      <td>87.330002</td>\n",
       "      <td>125.050003</td>\n",
       "      <td>44.199651</td>\n",
       "      <td>...</td>\n",
       "      <td>22.129999</td>\n",
       "      <td>24.700001</td>\n",
       "      <td>63.252712</td>\n",
       "      <td>27.200001</td>\n",
       "      <td>32.930000</td>\n",
       "      <td>121.349998</td>\n",
       "      <td>91.430000</td>\n",
       "      <td>41.770000</td>\n",
       "      <td>12.30</td>\n",
       "      <td>53.064997</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 984 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    A         AA        AAL         AAP        AAPL  \\\n",
       "Date                                                                  \n",
       "2017-03-28  53.307702  32.919998  42.599998  148.360001  143.208894   \n",
       "2017-03-29  52.988494  32.910000  41.959999  149.669998  143.527571   \n",
       "2017-03-30  53.188000  33.490002  42.540001  148.669998  143.338349   \n",
       "2017-03-31  52.869999  34.400002  42.299999  148.259995  143.069470   \n",
       "2017-04-03  53.200001  34.220001  42.450001  145.500000  143.109299   \n",
       "\n",
       "                  ABB       ABBV        ABC        ABMD        ABT    ...      \\\n",
       "Date                                                                  ...       \n",
       "2017-03-28  22.665505  65.330055  87.500000  125.050003  44.428253    ...       \n",
       "2017-03-29  22.588149  65.122117  89.510002  124.879997  44.398432    ...       \n",
       "2017-03-30  22.568809  64.815174  89.139999  125.000000  44.279163    ...       \n",
       "2017-03-31  22.626826  64.518133  88.500000  125.199997  44.140015    ...       \n",
       "2017-04-03  22.472113  64.389409  87.330002  125.050003  44.199651    ...       \n",
       "\n",
       "                 YNDX        YPF        YUM       YUMC       ZAYO         ZBH  \\\n",
       "Date                                                                            \n",
       "2017-03-28  22.430000  23.900000  64.168410  26.549999  32.900002  120.950002   \n",
       "2017-03-29  22.240000  23.850000  63.750376  26.610001  32.910000  121.370003   \n",
       "2017-03-30  22.059999  24.180000  63.720510  26.520000  32.810001  121.360001   \n",
       "2017-03-31  21.930000  24.280001  63.601077  27.200001  32.900002  122.110001   \n",
       "2017-04-03  22.129999  24.700001  63.252712  27.200001  32.930000  121.349998   \n",
       "\n",
       "                 ZBRA       ZION    ZTO        ZTS  \n",
       "Date                                                \n",
       "2017-03-28  88.580002  41.299999  12.88  54.053044  \n",
       "2017-03-29  88.830002  41.320000  13.27  53.713715  \n",
       "2017-03-30  89.559998  42.380001  13.01  53.364408  \n",
       "2017-03-31  91.250000  42.000000  13.09  53.264603  \n",
       "2017-04-03  91.430000  41.770000  12.30  53.064997  \n",
       "\n",
       "[5 rows x 984 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "128"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.isnull(df).sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BF.B     32\n",
       "BRK.A    32\n",
       "FCE.A    32\n",
       "RDS.A    32\n",
       "dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "to_remove = pd.isnull(df).sum()[pd.isnull(df).sum() > 0]\n",
    "to_remove"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['BF.B', 'BRK.A', 'FCE.A', 'RDS.A'], dtype='object')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "to_remove.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = df.drop(to_remove.index,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.isnull(df).sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(32, 980)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ARIMA\n",
    "\n",
    "In order to select best model we will use the [AIC](https://en.wikipedia.org/wiki/Akaike_information_criterion), i.e. we will take the model with __lowest AIC__. \n",
    "\n",
    "### How it works on 1 security "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from statsmodels.tsa.arima_model import ARIMA\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "## let's start with an example \n",
    "series = df[\"AAPL\"]\n",
    "series = series.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xl8lNXZ//HPlcm+kYQEwhYS9p0QIxBEi1YRV7RaK6i4\noGBdqk8fn/prfarWtnaxitYdBUEfBXfRFivWBdCwJewkbIEQspCNJQskJDPn90cGGkNIJpOZTGZy\nvV+veZHc21y3Y765c+5znyPGGJRSSvkuP08XoJRSyr006JVSysdp0CullI/ToFdKKR+nQa+UUj5O\ng14ppXycBr1SSvk4DXqllPJxGvRKKeXj/D1dAEBsbKxJTEz0dBlKKeVVMjMzy4wxca1t1ymCPjEx\nkYyMDE+XoZRSXkVEDjiynTbdKKWUj9OgV0opH6dBr5RSPk6DXimlfFyrQS8iC0WkRES2N1r2uIgU\niMhm++vyJvskiEiViDzkjqKVUko5zpEr+kXAtGaWzzPGJNtfy5uuAz5vb3FKKaXar9XulcaYVSKS\n6OgBReQaYB9Q7XxZSimlXKU9bfT3ichWe9NONICIhAEPA79zSXVKKaWa9c3OEoe3dTboXwYGAslA\nEfC0ffnvaGjSqWrtACIyR0QyRCSjtLTUyTKUUqrrqa238tinOxze3qknY40xxae+FpHXgH/Yv50A\nXC8ifwWiAJuI1BhjXmjmGPOB+QCpqak6Q7lSSjlo4Xe55B0+7vD2TgW9iPQyxhTZv70W2A5gjDm/\n0TaPA1XNhbxSSinnlFTW8OI3e7l4eA8WOLhPq0EvIkuAKUCsiOQDjwFTRCQZMEAuMNe5kpVSSrXF\n01/sprbeyiNXjHBd0BtjZjSzuNXjG2Med7AGpZRSDthecIz3Mg9y5+QkkmLDHN5Pn4xVSikvYIzh\nic+yiA4N5L6LBrdpXw16pZTyAp9vP8T63MP899QhdAsJaNO+GvRKKeVmxrSvY2FNnZUnl2czLD6C\nG89NaPP+GvRKKeVGq/eUkvzEl7yzLs/pYyz4bj/5R07w6JUjsPhJm/fXoFdKKTfJP3KcXyzZxPGT\n9fzm420s/G5/m49RUtHQnXLqiJ5MGhTrVB0a9Eop5Qa19VbufXsj9VbDP39xPtNGxvPEP7J48Zu9\nbTrOX7/YRb3V8MgVw52uRYNeKaXc4InPstiSf4y/3TCWIT0jeGHmOKYn9+apL3bx9IpdDrXbb80/\nygeZ+dw+OZH+3R3vTtlUp5gcXCmlfMmHmfm8vS6PuT8awKUj4wHwt/jxzA3JhARYeP7rvZw4aeWR\nK4Yj0nyb+6nulLHhgdx34aB21aNBr5RSLpRdVMEjn2xj4oAY/mfq0B+ss/gJT147muAAC69/t58T\ndVZ+P30Ufs3cYP3H1iIyDhzhzz8ZTURw27pTNqVBr5RSLnLsRB13/18m3UICeH5GCv6WM1vH/fyE\nx64aQXCAhVdW5lBTZ+Ov14/5QW+amjorf/58JyN6RfLT1H7trkuDXimlXMBmMzz0/hYKjpxg6ZyJ\nxEUEnXVbEeHhaUMJCbAw798NY9fM+1kyAfZfDPNX7aPg6AmevmGsU90pm9KgV0opF3h11T6+zCrm\n0StHkJoY0+r2IsIDFw8mJNCPJ5fvpKbOxos3jeNIdR0vf5vDZaPimTigu0tq06BXSql2St9bxlNf\n7OTKMb24/bzENu0754KBBAdYeHTZDu5cnEFkSABWY/jN5c53p2xKg14ppdrh0LEa7l+yiQFx4fzl\nujFn7UXTkllpiQT7W3j4o60YA/dMGUi/mFCX1ahBr5RSTjpZb+OetzOpqbPyys0phAU5H6k3nNuP\nsCB/PtlcwD3t7E7ZlAa9Uko56cnl2WzMO8oLM8cxqEdEu493xZheXDGmlwsq+yF9MlYppZywbHMB\ni9JzueO8JK4c09vT5bRIg14ppdqops7KIx9vJ7V/NL++fJiny2mVBr1SSrXR7uJKqmrrmT056XTf\n986s1QpFZKGIlIjI9kbLHheRAhHZbH9dbl9+iYhkisg2+78XubN4pZTyhKzCCgBG9I70cCWOceRX\n0SJgWjPL5xljku2v5fZlZcBVxpjRwK3AW64pUymlOo+sogrCg/zpF+26LpDu1GqvG2PMKhFJdORg\nxphNjb7dAQSLSJAxpta58pRSqvPJKqxgeK+IZgcj64za07h0n4hstTftRDez/jpgk4a8UsqX2GyG\n7KIKRvTyjmYbcD7oXwYGAslAEfB045UiMhL4CzD3bAcQkTkikiEiGaWlpU6WoZRSHSvv8HGqT1q9\npn0enAx6Y0yxMcZqjLEBrwHjT60Tkb7Ax8AsY0xOC8eYb4xJNcakxsXFOVOGUkp1uKwi+43YXt08\nXInjnAp6EWn86Na1wHb78ijgn8CvjTHft788pZTqXLIKK7D4CYN7hnu6FIe1ejNWRJYAU4BYEckH\nHgOmiEgyYIBc/tNEcx8wCPitiPzWvmyqMabExXUrpZRHZBVVMCgunOAAi6dLcZgjvW5mNLN4wVm2\n/QPwh/YWpZRSnVVWYQVpA10zTnxH6fyPdCmlVCdRXlXLoYoar+pxAxr0SinlsOyiSsB7nog9RYNe\nKaUclFV0DIDhekWvlFK+Kauwgl7dgokJC/R0KW2iQa+UUg7K8rInYk/RoFdKKQfU1FnJKa32uvZ5\n0KBXSimH7C6uxGozekWvlFK+ytvGoG9Mg14ppRzgbWPQN6ZBr5RSDvC2Megb06BXSqlWeOMY9I1p\n0CulVCu8cQz6xjTolVKqFd44Bn1jGvRKKdUKbxyDvjENeqWUaoU3jkHfmAa9Ukq1Iquwwmvb50GD\nXimlWuStY9A3pkGvlFIt8NYx6BvToFdKqRZ46xj0jWnQK6VUC7x1DPrGWg16EVkoIiUisr3RssdF\npEBENttflzda92sR2Ssiu0TkUncVrpRSHcFbx6BvzJEr+kXAtGaWzzPGJNtfywFEZARwIzDSvs9L\nIuKd/ZGUUl2eN49B31irQW+MWQUcdvB404GlxphaY8x+YC8wvh31KaWUx3jzGPSNtaeN/j4R2Wpv\n2om2L+sDHGy0Tb59mVJKeR1vHoO+MWeD/mVgIJAMFAFP25c3N36nae4AIjJHRDJEJKO0tNTJMpRS\nyn28eQz6xpwKemNMsTHGaoyxAa/xn+aZfKBfo037AoVnOcZ8Y0yqMSY1Li7OmTKUUsqtvHkM+sac\nCnoR6dXo22uBUz1yPgVuFJEgEUkCBgPr21eiUkp1PG8fg74x/9Y2EJElwBQgVkTygceAKSKSTEOz\nTC4wF8AYs0NE3gOygHrgXmOM1T2lK6WU+xzw8jHoG2s16I0xM5pZvKCF7f8I/LE9RSmllKedvhHr\npWPQN6ZPxiqlVDOyio559Rj0jWnQK6VUM7IKvXsM+sY06JVSqhlZRd49Bn1jGvRKKdVEWVUtxRW1\nPtHjBjTolVLqDNlFvvFE7Cka9Eop1cSpHjfePAZ9Yxr0SinVRFaR949B35gGvVJKNZFV6BtPxJ6i\nQa+UUo00jEFf5TPt86BBr5RSP7DrUCU2g17RK6WUr8rysR43oEGvlFI/kFXoG2PQN6ZBr5RSjZya\nDNzbx6BvTINeKaXsTo9B70PNNqBBr5RSpx04fJzjJ60+dSMWNOiVUuo0X5kMvCkNeqWUsssqOoa/\nnzCoh/ePQd+YBr1SStllFVYwqIdvjEHfmAa9UkoBdVYbmQeOMKav908d2JQGvVJKAev2Haaipp5L\nRsR7uhSXcyjoRWShiJSIyPZm1j0kIkZEYu3fdxORz0Rki4jsEJHbXV20Ukq52oqsQ4QEWDh/cKyn\nS3E5R6/oFwHTmi4UkX7AJUBeo8X3AlnGmLHAFOBpEfGNsT6VUj7JGMOKHcVcMCTW59rnwcGgN8as\nAg43s2oe8CvANN4ciBARAcLt+9W3s06llHKbbQXHOFRRw1QfbLaBdrTRi8jVQIExZkuTVS8Aw4FC\nYBvwgDHG1sz+c0QkQ0QySktLnS1DKaXabcWOYix+wkXDeni6FLdwKuhFJBR4BHi0mdWXApuB3kAy\n8IKInPH0gTFmvjEm1RiTGhcX50wZSinlEiuyDjE+MYZoH5lRqilnr+gHAknAFhHJBfoCG0UkHrgd\n+Mg02AvsB4a5olillHK1/WXV7C6uYurInp4uxW38ndnJGLMNOP03jj3sU40xZSKSB/wYWC0iPYGh\nwD4X1KqUUi73ZdYhAC4Z4btB72j3yiXAGmCoiOSLyOwWNv89MElEtgFfAQ8bY8raX6pSSrneih3F\njOwdSV8fGn++KYeu6I0xM1pZn9jo60JgavvKUkop9yutrCUz7wgP/niIp0txK30yVinVZX2VXYwx\n+HT7PGjQK6W6sBVZxfSLCWFYfISnS3ErDXqlVJdUVVvPd3vLmDoinobnO32XBr1SqktauauUk/U2\npvpwb5tTNOiVUl3SiqxDxIQFck7/aE+X4nYa9EqpLudkvY2vd5bw42E98Lf4fgz6/hkqpVQT6/aX\nU1lTz9SRvjmIWVMa9EqpLmfFjmKfHXu+ORr0SqkuxWYzfJnlu2PPN0eDXinVKWw+eJSLn1nJS9/u\npbbe6rb38fWx55ujQa+U8rjiihrmvJlB0dET/PVfu5j27Gq+2VXilvdakXXIp8eeb44GvVLKo2rq\nrMx5K5Oq2no+vGcSb94xHhG4/Y0N3Ll4AwfKq136fit2FPv02PPN0aBXSnmMMYbffLSNLQeP8swN\nyQyLj+SCIXH864EL+PVlw1iTU84l81bxzIpdnDjZ/uacfaVV7Cmp4lIfH9umKQ16pZTHvL56Px9t\nKuC/Lh7CtFH/aTMP9Pdj7o8G8vVDU7h8VDx//3ovFz+zks+3FWGMaeGILfsyqxiAS7pIt8pTNOiV\nUh6xcncpf/o8m8tHx3P/RYOa3aZnZDDP3jiO9+amERHsz8/f3sjNC9axp7jSqfdckVXMqD6R9IkK\naU/pXkeDXinV4faVVnHfOxsZGh/J3346Fj+/lgcVG58Uwz/un8wT00eyLf8Ylz23mj/+M6tNzTkl\nlTVszDvSpXrbnKJBr5TqUBU1ddz5ZgYBFj9em3UOoYGOzWjqb/FjVloi3zw0hevP6ctrq/dz5fOr\n2ZZ/zKH9v8ou6RJjzzdHg14p1WGsNsMDSzaRV36cl29KcWr6vu7hQfz5ujH83+wJVNdaufal73nh\n6z3UW20t7rdixyESYkIZ2tO3x55vjga9UqrD/PWLnXyzq5THrx7JhAHd23WsyYNj+deD5zNtVDx/\nW7Gbn81fe9aumFW19Xy/t5ypI3r6/NjzzWk16EVkoYiUiMj2ZtY9JCJGRGIbLZsiIptFZIeIrHR1\nwUop7/TJpgJeXbmPmyYkcPPE/i45ZlRoIC/MTOG5G5PZXVzJ5c+t5t0NeWf0zFm5q5STVluXGcSs\nKUeu6BcB05ouFJF+wCVAXqNlUcBLwNXGmJHAT11TplLKm23NP8rDH25lfFIMj1010uXHn57chy8e\nvIAxfaN4+MNtzHkrk7Kq2tPru9LY881pNeiNMauAw82smgf8Cmj8q3Mm8JExJs++r3ueYVZKeY2S\nihrmvJlJbHgQL9+UQqC/e1qMe0eF8PadE/jfK4azcncp055dxVfZxafHnr94eA8srfTu8VVO/RcX\nkauBAmPMliarhgDRIvKtiGSKyKx2V6iU8lrGGO55eyPHTtTx2qxUuocHufX9/PyEO88fwGf3TSY2\nPIjZizO47Y31DWPPd8Fulae0OehFJBR4BHi0mdX+wDnAFcClwG9FZMhZjjNHRDJEJKO0tLStZSil\nvMCekioyDhzh4WlDGdE7ssPed2h8BMvuO4+5Fwxgzb5yQgIsTO4iY883x7EOrD80EEgCttjvXvcF\nNorIeCAfKDPGVAPVIrIKGAvsbnoQY8x8YD5Aamqq8880K6U6re/3lgFwsQcm4A7yt/Dry4dz6ah4\nqmvru8zY881pc9AbY7YBp8f3FJFcINUYUyYiy4AXRMQfCAQm0NCWr5TqgtJzykmICXWqv7yrpCR0\nzRuwjTnSvXIJsAYYKiL5IjL7bNsaY7KBfwFbgfXA68aYM7plKqV8X73Vxtp95Zw3qH395VX7tXpF\nb4yZ0cr6xCbfPwU81b6ylFLebkdhBZU19aQN7Lpt452FPhmrlHKL9JxyANLa+QSsaj8NeqWUW6Tn\nlDG0ZwRxEe7tUqlap0GvlHK52norG3IPkzZQr+Y7Aw16pZTLbco7Sk2djfMGaft8Z6BBr5RyufSc\ncvykYcIQ5Xka9Eopl1uTU8boPt3oFhLg6VIUGvRKKRerrq1nU95RJmmzTaehQa+UcqkNuYeptxkm\n6Y3YTkODXinlUmtyygm0+JHaX9vnOwsNeqWUS32fU8a4hChCArvuIGKdjQa9Uspljh4/yY7CCibp\nsAediga9Uspl1u4rxxh0ILNORoNeKeUy6TnlhAZaGNM3ytOlqEY06JVSLpOeU865iTFumxdWOUc/\nDaWUSxRX1LC3pEqbbTohDXqllEussQ9LrDdiOx8NeqWUS6TnlNEtJIDhvTpuEnDlGA16pVS7GWP4\nfm85aQO6Y/ETT5ejmtCgV0q128HDJyg4eoJJ2j7fKWnQK6Xa7fucMkDb5zsrh4JeRBaKSImIbG9m\n3UMiYkQktsnyc0XEKiLXu6pYb/JhZj4X/e1bthcc83QpSrldek45PSKCGBgX5ulSVDMcvaJfBExr\nulBE+gGXAHlNlluAvwBftLM+r7S3pIpHPtnGvrJqbnp9HdvyNeyV7zLGsCanjEkDuyOi7fOdkUNB\nb4xZBRxuZtU84FeAabL8fuBDoKRd1Xmhk/U2Hnx3EyEBFj78eRrhQf7c9Ppathw86unSlHKL3cVV\nlFWd1PHnOzGn2+hF5GqgwBizpcnyPsC1wCut7D9HRDJEJKO0tNTZMjqdZ77czfaCCv583RjO6R/D\nu3Mn0i00gJtfX8emvCOeLk/5gMqaOuavyuFI9UlPlwI0dKsEdPz5TsypoBeRUOAR4NFmVj8LPGyM\nsbZ0DGPMfGNMqjEmNS4uzpkyOp30nDJeXZXDjPEJXDoyHoC+0aG8OyeNmPBAblmwnswDzf1hpJRj\n6q027n1nE08u38kdizdw4mSLP2YdIj2nnISYUPpGh3q6FHUWzl7RDwSSgC0ikgv0BTaKSDyQCiy1\nL78eeElErnFBrZ3aseN1/Pd7W0jqHsZvrxz+g3W9o0JYOmcicRFBzFqwng25Gvaq7YwxPPbpDlbt\nLuWn5/Rl88Gj3L9kI/VWm8dqqrfaWLuvXIc96OScCnpjzDZjTA9jTKIxJhHIB1KMMYeMMUmNln8A\n3GOM+cR1JXc+xhh+8/E2SitrefbGZEID/c/Yple3hrDvGRnMrQvXs25fuQcqVd5swXf7eXtdHnMv\nGMBTPx3L764eyb+zS/jtsu0Y0/Q2WcfYUVhBZU09adqtslNztHvlEmANMFRE8kVktnvL8i4fbizg\nn9uK+OXUIS0Oz9ozMpilcybSq1swt72x4fTYIEq15susYv64PJtpI+N5eNowAGalJXLPlIEsWX+Q\nv3+11yN1neo/nzZAr+g7M0d73cwwxvQyxgQYY/oaYxY0WZ9ojClrZr/bjDEfuKrYzuhAeTWPLdvO\nhKQY5l4wsNXte0QGs3ROGn2jQ7h90Xq+33vGfzalfmB7wTF+sWQTo/t0Y97PkvFrNMTA/1w6lOtS\n+jLv37tZuj6vhaO4x5qccob2jCAuIqjD31s5zmuejD14+Dh/+jyb+5ds4q01uewprvTYn6un1Ftt\nPPjuZix+wryfJTs8xkdcRBBL5kykf0wYdyzawKrdvtPrSLlW0bETzF68gejQAF6flXrGPKwiwp+v\nG82PhsTxyCfb+Sq7uMNqq623siH3MGna26bTO7MxuROx2Qzf7S3jzTW5fLWzBD8RYsMD+WxLIQDd\nwwKZOKA7EwfEMHFAdwb1CO/QBzae/3ovm/KO8sLMcfSOCmnTvrHhDWE/87W13PlmBi/NTOHiET3d\nVKnyRtW19cxelEF1rZUPfp5Gj8jgZrcLsPjx0k0p3Dh/Lfe+s5F37ppISkK02+vblHeUmjob52n/\n+U6vUwZ9RU0dH2Tk89baA+wvqyY2PJD7LhzEzAkJxEcGc/DwCdbuKz/9+ue2IqBjgz/zwGGe/3oP\nP0npw5Vjejt1jJiwQJbcNZGbF6zjzjczmDgghnumDOL8wbH6hGEXZ7UZHli6iZ2HKlhw27kMi295\n6N+wIH8W3nYu17+SzuxFG/jg55MYGBfu1hrTc8rxExifFOPW91HtJ55u/gBITU01GRkZ7DpUyZtr\ncvl4UwHHT1oZlxDFrWmJXDY6niB/S7P7GmPOCP7CYzUA9IkKYXpyb64Z14chPSNcVm9lTR2XPbca\nEVj+i/OJCA5o1/GOn6znnXV5vLZ6H8UVtYzu0417LxzI1BHxP2iPVV3HE59lsfD7/fx++khuSUt0\neL/csmquezmdkEALH/180ln/CnCF619Op85qY9l9k932HqplIpJpjEltdbvOEPSDR441Kb94hXX7\nDxPo78f0sb2ZlZbI6L7d2nwsYwz5R06QnlPG8m2H+G5vGVabYXivSK5J7s3Vyb3p1a1tzSxN/fLd\nzSzbUsh7c9M4p7/r/kSurbfy8cYCXlmZQ275cQbGhXH3jwZyzbg+BFi85naKaqe31uTy22U7uP28\nRB67amSb999y8Cg3zl9LUmwY786d6PCFiM1mMODQvabq2nrG/m4Fd10w4HQvINXxvCrog3oNNqkP\nvMotaf25IbUfMWGBLjt2aWUt/9hayCebC9ly8CgiMCEphmuS+3DZqF50C23b1finWwr5xZJNPHjx\nYB68eIjL6mzMajMs31bES9/mkF1UQZ+oEO46P4mfnZtwxs045Vu+3VXCHYs2cOHQHsyfler0JB7f\n7CrhzsUZpA3ozsLbzj09WbcxhtLKWvaXVZNbXs3+suPk2r/OLa/GZoPBPcMZ3ivS/opgRK9IokJ/\n+DP57a4SbntjA2/NHs/5g33jyXZv5FVBP2TUWJO9dbPbZ6bZX1bNss0FLNtcyP6yagItfkwZGsf0\n5D7EhgdSfbKeypp6qmutVNXWUVVrpaqmnuraeqrsr8wDRxjSM5z35qbh7+arbGMM3+4u5aVv9rIh\n9wjdwwK5Y3ISN0/sT7eQ9jUXqc5n56EKrn95DQkxobx/dxphQe27hfZ+xkH+54OtnDeoO1Ehgewv\nq+ZAeTXVjYZNCLAI/WJCSeoeRmJsGP5+QlZRBdlFlZRV1Z7erle3YEacDv9IVu4u4ZNNhWx5bKpe\nfHiQVwX9qTb6jmKMYWv+MT7ZXMBnW4p+8D90U2GBFsKD/QkL8iciyJ+4iGAeu2oE/WI6dlyPDbmH\neembvXyzq5SQAAvXjOvDrLT+Oj+nj6iurWfqvFXU22x8cu957W5ePOXVlTk899UeekQEkRgbRmL3\nMJJiG0I9qXsYvaOCz3rBUlpZS3ZRBdlFFfbwryCntBqrrSEzJiTF8O7cNJfUqZyjQe+gequNjXlH\nOVlvIzzYn/AgC+FBAYQFWQgL9O90N0OzCitYnJ7LJ5sLqK23MT4phlvTEpk6sqe243uxeV/u5rmv\n9vDB3WmkJrq2F4sxxmW9uGrqrOwtqWLnoUqS+3VjUA/XdXJQbadB7+OOHj/J+xn5vLk2l4OHT9Az\nMoiZ4/szY0I/ekS4r6eFcr3iihqmPPUtFw3vwYszUzxdjvIiGvRdhNVmWLm7hMXpB1i5u5QAi3DZ\nqF7cOqk/KQnR2h/fCzz8wVY+2pTPV7+cQkJ3HepXOc7RoO+UD0wpx1n8hIuG9eSiYT3ZX1bNW2sO\n8H7mQT7dUsjI3pFMHRFPSv8okvtFtbu/v3K97KIK3ss8yOzzkjTkldvoFb0POn6ynk82FbJkfR7b\nC49hDIjAkB4RpPSPJiUhipT+0QyIDdMrfg+btXA9Ww4eZeX/TDmjC6NSrdEr+i4sNNCfmRMSmDkh\ngYqaOrYcPMrGA0fZmHeEf25t+AUAEB0awLiEhuCfOKC7y28Cqpat3F3Kqt2l/O8VwzXklVtp0Pu4\nyOAAzh8cd/qhFpvNkFNaxca8I6fD/+udDXO4Pz9jHFeNdW7cHtU2VpvhT8uzSYgJ5Za0/p4uR/k4\nDfouxs9PGNwzgsE9I/jZuQlAwzSIN7y6hue+2sPlo3u5/cE1BR9m5rPzUCUvzkw56zhOSrmKdrxW\ndAsN4P4fD2JvSRXL7SOBKvc5frKep7/cxbiEKC4fHe/pclQXoEGvALhsVC8G9Qjn+a/3YLN5/ga9\nL3t99X6KK2r53yuG681w1SE06BXQ0E3z/osGsbu4ii92HPJ0OT6rpLKGV1bmcPnoeM7prze/Vcdo\nNehFZKGIlIjI9mbWPSQiRkRi7d/fJCJb7a90ERnrjqKVe1w5pjcDYsN47iu9qneXeV/uoc5q41eX\n6tC+quM4ckW/CJjWdKGI9AMuARrPSLwf+JExZgzwe2C+C2pUHcTiJ9x30SB2Hqrkyw6ce7Sr2F1c\nybsb8rhlYiKJsWGeLkd1Ia0GvTFmFXC4mVXzgF8BptG26caYI/Zv1wJ9XVGk6jhXj+1NYvdQ/v7V\nHo9Pvu5r/rQ8m/Agf+6/aJCnS1FdjFNt9CJyNVBgjNnSwmazgc+dqkp5jL/Fj3svHMSOworT/etV\n+323p4xvdpVy/0WDiXbhxDpKOaLNQS8iocAjwKMtbHMhDUH/cAvbzBGRDBHJKC0tbWsZyo2uGdeH\nfjEhPKdX9S5htRn+uDybvtEhzJqkD0epjufMFf1AIAnYIiK5NDTPbBSReAARGQO8Dkw3xpSf7SDG\nmPnGmFRjTGpcnE5F1pkEWPy4d8ogtuYf49vd+ku4qbfWHuDal77n8U938OmWQvKPHG/xF+LHmwrI\nLqrg4WnD9OEo5RFtfjLWGLMN6HHqe3vYpxpjykQkAfgIuMUYs9tlVaoO95OUvjz/9V6e+/cepgyJ\n0/7edjV1Vp5ZsQuLn7CzqJJF6bkA9IwMIiUhuuHVP5pRfSIJ8rdw4qSVv32xi+R+UVw5ppdni1dd\nVqtBLyLjrlcjAAALs0lEQVRLgClArIjkA48ZYxacZfNHge7AS/ZgqHdkZDXV+QT6+3HPhQN55OPt\nrN5TxgVD9K8ugGWbCzhyvI4ld00kNTGanUWVDeMG5R0h88ARPt/e8AxCoMWPkX0iCQ/y51BFDc/P\nHKe/LJXH6DDF6qxq661MeepbekeF8MHdaV0+qIwxXPbcagA+f+D8Zv97lFTWsPHAUTbZg39rwTEu\nHRnP8zPGdXS5qgvQYYpVuwX5W/j5lIE8umwHa3LKmTQo1tMledTafYfZeaiSv1w3+qy/9HpEBDNt\nVDzTRjWMYVNvtekgccrjdAgE1aIbUvvRMzKI577a4+lSPG5R+n6iQwOYntzH4X38LX5d/i8h5Xka\n9KpFwQEW7v7RQNbtP8zafWftROXzDh4+zpdZxcwYn0BwgPacUd5Fg161asb4BOIigvh7F76qf2vt\nAUREJwlRXkmDXrUqOMDC3AsGkJ5Tzobc5kbD6FiHq09i7cBB146frGfp+jymjYqnV7eQDntfpVxF\ng1455KYJ/YkND2z1qr7eamN7wTHeWpPLf727mSeXZ1Nntbmsjm92lTDxya+46fW1HKk+6bLjtuSj\njQVU1NRzx3mJHfJ+Srma9rpRDgkJtHDX+QP40+c7yTxwhHP6RwNwpPrk6X7kGw8cZUv+UY6ftALQ\nPSyQ8uqT5JZV88LMFAL923dd8d2eMua+lUnf6BA25h1l+ovfs+DWVAb3jGj3+Z2NMYZF6bmM6duN\nlIRot72PUu6kQa8cdvPE/ry6ah9PfLaDQT0i2JR3hH1l1UDDEMcjekXy03P6ktK/4QnRvtEhLE7P\n5fHPsrjn7UxevMn5+VHX7ivnzjc3MCA2jKVzJrKvrJo5b2byk5fS+fvMcVw4tEfrB3HC6j1l7C2p\n4pkbxmrvGeW19IEp1SbzV+Xw5PKdxIQFkpIQdTrUx/TtRmhg89cNb63J5bfLdnDRsB68dFNKm3ut\nZOQeZtbC9fSJCmHpnIl0Dw8CoPDoCe56M4Psogp+c/lwZk9OcnkY37FoA1vzj/H9/7tQx6lRnY4+\nMKXc4s7JA5ie3IceEUEOh+otaYlY/Pz4zcfbmPtWJq/eco7DYb/54FFue2MD8ZHBvH3XhNMhD9A7\nKoT3707jv9/bwh/+mc2uQ5X84dpRLgvk/WXVfL2zhAd+PFhDXnk1vRmr2sTPT+gZGdzmK+eZExL4\n809Gs2pPKXe9mUFNnbXVfbYXHOOWBevoHh7IO3dNpEdE8BnbhAb68+LMFH7x48G8n5nPza+vo6yq\ntk21nc3i9FwCLMJNExNccjylPEWDXnWYG8cn8JfrxvDd3jJmL97AiZNnD/vsogpuXrCOyOAA3rlr\nIvHdzgz5U/z8hF9eMoTnZ4xja/4xpr/wPdlFFe2qtbKmjg8y87lyTO9mf8Eo5U006FWHuiG1H3+7\nfizpOeXcsWgDx0/Wn7HNnuJKbn59HSEBFpbcNZE+UY71Xb9qbG/evzuNepuN615OZ8WOQ07X+UFm\nPlW19dw2KdHpYyjVWWjQqw533Tl9mXdDMuv2l3P7Gxuorv1P2O8rrWLm6+uw+Anv3DWRhO6hbTr2\nmL5RfHrfZAb3CGfu/2Xy4jd72zxLls1mWJyeS0pCFGP7RbVpX6U6Iw165RHXjOvDvJ8lsyH3MLe/\nsYGq2noOlFcz87V1GGN4564JJMWGOXXsnpHBvDs3javG9OapL3Zx2xsbOHSsxuH9v91dQm75cW4/\nL8mp91eqs9FeN8pjpif3weInPLB0M7csWEdJRS219VaWzJnIoB7tewgqOMDCczcmk5oYzZPLs7n0\n2VU8MX0kV4/t3eqN5De+zyU+Mvj0UMNKeTu9olcedeWY3rwwYxzb8o9RWVPHW7MnMCw+0iXHFhFm\npSXy+QMXMDAujAeWbubedzZyuIWhE/YUV7J6Txm3pPUnwKI/Hso36BW98rjLRvfiw6gQIkMCnG6u\naUlSbBjv3z2JV1flMO/L3azff4S/XDeaHw/veca2i9JzCfT348Zz+7m8DqU8RS9ZVKcwtl+UW0L+\nFIufcM+UQSy7dzKx4YHMXpzBrz7YQmVN3eltjh2v46ONBVyT3PsHD2Yp5e006FWXMqJ3JMvuO497\npgzkg8x8pj27mjU5DROqvJuRx4k6K7dN0puwyrc4FPQislBESkRkezPrHhIRIyKx9u9FRP4uIntF\nZKuIpLi6aKXaI8jfwq+mDeP9uycRYBFmvLaW3322g8XpB5iQFMOI3q65R6BUZ+HoFf0iYFrThSLS\nD7gEyGu0+DJgsP01B3i5fSUq5R7n9I9m+QPnMyutP298n0vB0RPcrmPOKx/k0M1YY8wqEUlsZtU8\n4FfAskbLpgNvmoanVNaKSJSI9DLGFLW3WKVcLTTQnyemj2LqiHjW7Cvj4mZu0Crl7ZzudSMiVwMF\nxpgtTfol9wEONvo+377sB0EvInNouOInIUEHjVKeNXlwLJMHx3q6DKXcwqmbsSISCjwCPNrc6maW\nnfEMujFmvjEm1RiTGhcX50wZSimlHODsFf1AIAk4dTXfF9goIuNpuIJv3Am5L1DYniKVUko5z6kr\nemPMNmNMD2NMojEmkYZwTzHGHAI+BWbZe99MBI5p+7xSSnmOo90rlwBrgKEiki8is1vYfDmwD9gL\nvAbc0+4qlVJKOc3RXjczWlmf2OhrA9zbvrKUUkq5ij4Zq5RSPk6DXimlfJwGvVJK+Thp6zRrbilC\npBLY5ek63CAWKPN0EW6g5+Vd9Ly8S1vOq78xptUHkTrLePS7jDGpni7C1UQkQ8/Le+h5eRc9L8dp\n041SSvk4DXqllPJxnSXo53u6ADfR8/Iuel7eRc/LQZ3iZqxSSin36SxX9EoppdzE40EvItNEZJd9\n6sH/5+l6XEVEckVkm4hsFpEMT9fjrOamkRSRGBH5UkT22P+N9mSNzjjLeT0uIgX2z2yziFzuyRqd\nISL9ROQbEckWkR0i8oB9uVd/Zi2cl1d/ZiISLCLrRWSL/bx+Z1+eJCLr7J/XuyIS2K738WTTjYhY\ngN00TEeYD2wAZhhjsjxWlIuISC6Qaozx6n6+InIBUEXDrGGj7Mv+Chw2xvzZ/ss52hjzsCfrbKuz\nnNfjQJUx5m+erK09RKQX0MsYs1FEIoBM4BrgNrz4M2vhvG7Aiz8zaRjnPcwYUyUiAcB3wAPAL4GP\njDFLReQVYIsxxulpWT19RT8e2GuM2WeMOQkspWEqQtVJGGNWAYebLJ4OLLZ/vZiGHzivcpbz8nrG\nmCJjzEb715VANg0zvHn1Z9bCeXk106DK/m2A/WWAi4AP7Mvb/Xl5OujPNu2gLzDAChHJtE+b6Et6\nnppjwP5vDw/X40r3ichWe9OOVzVvNGWf53kcsA4f+syanBd4+WcmIhYR2QyUAF8COcBRY0y9fZN2\n56Kng96haQe91HnGmBTgMuBee1OB6txepmH2tGQa5jh+2rPlOE9EwoEPgQeNMRWersdVmjkvr//M\njDFWY0wyDbPxjQeGN7dZe97D00Hvs9MOGmMK7f+WAB/T8AH6imJ7m+mpttMSD9fjEsaYYvsPnY2G\nSXO88jOzt/V+CLxtjPnIvtjrP7PmzstXPjMAY8xR4FtgIhAlIqeGqGl3Lno66DcAg+13mAOBG2mY\nitCriUiY/YYRIhIGTAW2t7yXV/kUuNX+9a3AMg/W4jKngtDuWrzwM7Pf3FsAZBtjnmm0yqs/s7Od\nl7d/ZiISJyJR9q9DgItpuP/wDXC9fbN2f14ef2DK3h3qWcACLDTG/NGjBbmAiAyg4SoeGgaOe8db\nz8s+jeQUGkbUKwYeAz4B3gMSgDzgp8YYr7qxeZbzmkJDE4ABcoG53jbfsYhMBlYD2wCbffFvaGjP\n9trPrIXzmoEXf2YiMoaGm60WGi683zPGPGHPkKVADLAJuNkYU+v0+3g66JVSSrmXp5tulFJKuZkG\nvVJK+TgNeqWU8nEa9Eop5eM06JVSysdp0CullI/ToFdKKR+nQa+UUj7u/wPI3nb9t9f5ywAAAABJ\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1ff48c81978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.Series(series).plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tr_size: 21 xval_size: 11\n"
     ]
    }
   ],
   "source": [
    "p_tr = 0.66\n",
    "\n",
    "tr_size = int(len(series) * p_tr)\n",
    "train, test = series[0:tr_size], series[tr_size:len(series)]\n",
    "print(\"tr_size:\",tr_size,\"xval_size:\",(len(series)-tr_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                              ARMA Model Results                              \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   No. Observations:                   21\n",
      "Model:                     ARMA(5, 0)   Log Likelihood                 -21.813\n",
      "Method:                       css-mle   S.D. of innovations              0.655\n",
      "Date:                Fri, 12 May 2017   AIC                             57.625\n",
      "Time:                        17:23:40   BIC                             64.937\n",
      "Sample:                             0   HQIC                            59.212\n",
      "                                                                              \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "const        142.4805      0.394    361.613      0.000     141.708     143.253\n",
      "ar.L1.y        0.5546      0.219      2.533      0.023       0.125       0.984\n",
      "ar.L2.y        0.3659      0.264      1.384      0.186      -0.152       0.884\n",
      "ar.L3.y        0.1509      0.265      0.568      0.578      -0.369       0.671\n",
      "ar.L4.y       -0.3398      0.252     -1.348      0.198      -0.834       0.154\n",
      "ar.L5.y       -0.1239      0.232     -0.534      0.601      -0.579       0.331\n",
      "                                    Roots                                    \n",
      "=============================================================================\n",
      "                 Real           Imaginary           Modulus         Frequency\n",
      "-----------------------------------------------------------------------------\n",
      "AR.1            1.0318           -0.3978j            1.1058           -0.0586\n",
      "AR.2            1.0318           +0.3978j            1.1058            0.0586\n",
      "AR.3           -0.8675           -1.1818j            1.4660           -0.3508\n",
      "AR.4           -0.8675           +1.1818j            1.4660            0.3508\n",
      "AR.5           -3.0706           -0.0000j            3.0706           -0.5000\n",
      "-----------------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAD8CAYAAACcjGjIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1ff4a5d5208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAD8CAYAAACYebj1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n1e7/A/KbBwBKUmueAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1ff4a5c3358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               0\n",
      "count  21.000000\n",
      "mean    0.017688\n",
      "std     0.685996\n",
      "min    -1.199004\n",
      "25%    -0.372014\n",
      "50%     0.073872\n",
      "75%     0.495655\n",
      "max     1.446828\n"
     ]
    }
   ],
   "source": [
    "from pandas import datetime\n",
    "from pandas import DataFrame\n",
    "from statsmodels.tsa.arima_model import ARIMA\n",
    "from matplotlib import pyplot\n",
    "\n",
    "# fit model\n",
    "model = ARIMA(train, order=(5,0,0))\n",
    "model_fit = model.fit(disp=0)\n",
    "print(model_fit.summary())\n",
    "# plot residual errors\n",
    "residuals = DataFrame(model_fit.resid)\n",
    "residuals.plot()\n",
    "pyplot.show()\n",
    "residuals.plot(kind='kde')\n",
    "pyplot.show()\n",
    "print(residuals.describe())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "57.625096989071324"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_fit.aic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from numpy.linalg import LinAlgError\n",
    "\n",
    "def get_best_arima(sr,maxord = [6,6,6]):\n",
    "    best_aic = 1000000\n",
    "    best_ord = maxord\n",
    "    best_model = None \n",
    "    for p in range(maxord[0]):\n",
    "        for q in range(maxord[1]):\n",
    "            for d in range(maxord[2]):\n",
    "                try: \n",
    "                    model = ARIMA(sr, order=(p,q,d))\n",
    "                    model_fit = model.fit(disp=0)\n",
    "                    if (best_aic > model_fit.aic):\n",
    "                        best_aic = model_fit.aic\n",
    "                        best_ord = [p,q,d]\n",
    "                        best_model = model_fit \n",
    "                except: \n",
    "                    pass\n",
    "    return best_aic, best_ord , best_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# supress warnings \n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best_aic: 49.735119320259955\n",
      "best_ord: [2, 0, 2]\n"
     ]
    }
   ],
   "source": [
    "best_aic, best_ord , best_model = get_best_arima(train)\n",
    "print(\"best_aic:\",best_aic)\n",
    "print(\"best_ord:\",best_ord)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def predict_arima(train,steps,best_order,verbose=True,do_plot=True):\n",
    "    history = [x for x in train]\n",
    "    predictions = list()\n",
    "    yhat = train[len(train)-1]\n",
    "    for t in range(steps):\n",
    "        model = ARIMA(history, order=(best_order[0],best_order[1],best_order[2]))\n",
    "        try:\n",
    "            model_fit = model.fit(disp=0)\n",
    "            output = model_fit.forecast()\n",
    "        except:\n",
    "            pass\n",
    "        if not(np.isnan(output[0] )) and not(np.isinf(output[0])): \n",
    "            yhat = np.asscalar(output[0])\n",
    "        predictions.append(yhat)\n",
    "        history.append(yhat)\n",
    "    return predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 35.0570845287\n"
     ]
    }
   ],
   "source": [
    "predictions = predict_arima(train,len(test),best_ord,verbose=True,do_plot=True)\n",
    "MSE = mean_squared_error(test, predictions)\n",
    "print(\"MSE:\",MSE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nnAa9Ukq5OA16pZRycf8Ph28fMF+SNAoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1ff4a6a3828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pyplot.plot(test , label=\"test data\")    \n",
    "pyplot.plot(predictions, color='red' , label=\"prediction\")\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def predict_arima_worse(train,steps,best_order,verbose=True,do_plot=True):\n",
    "    history = [x for x in train]\n",
    "    predictions = list()\n",
    "    model = ARIMA(history, order=(best_order[0],best_order[1],best_order[2]))\n",
    "    try:\n",
    "        model_fit = model.fit(disp=0)\n",
    "        output = model_fit.forecast(steps=steps)\n",
    "    except:\n",
    "        pass\n",
    "    yhat = history[len(history)-1]\n",
    "    for i in range(steps): \n",
    "        if not(np.isnan(output[0][i] )) and not(np.isinf(output[0][i])): \n",
    "            yhat = np.asscalar(output[0][i] )\n",
    "        predictions.append(yhat)\n",
    "    return predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 59.4889326175\n"
     ]
    }
   ],
   "source": [
    "predictions = predict_arima_worse(train,len(test),best_ord,verbose=True,do_plot=True)\n",
    "MSE = mean_squared_error(test, predictions)\n",
    "print(\"MSE:\",MSE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlFJu7v8DoMZ0sKL6uYYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1ff4a6982e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pyplot.plot(test , label=\"test data\")    \n",
    "pyplot.plot(predictions, color='red' , label=\"prediction\")\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Massive Predictions \n",
    "\n",
    "Let's do a simple experiments with above defined securities, i.e. for each security\n",
    "* let's fit best ARIMA model on train period (oldest 60%)  \n",
    "* let's predict stock price/cumulative return on cross-validation period (%20 in the middle)\n",
    "* let's measure MSE/R2 for price/cumulative return on cross-validation period\n",
    "* let's predict stock price/cumulative return on test period (most recent %20)\n",
    "* let's measure MSE/R2 for price/cumulative return on test period\n",
    "\n",
    "If we considered ARIMA models accurate then we should pick the securities with highest predicted cumulative return. Unfortunately, that would be a huge mistake. Why? Let's consider, for example in the cross validation period, predicted cumulative return vs. actual cumulative return. We would find out huge surprises. \n",
    "\n",
    "Another technique for model selection is using indicator like [AIC](https://en.wikipedia.org/wiki/Akaike_information_criterion) or [BIC](https://en.wikipedia.org/wiki/Bayesian_information_criterion) and actually in the below procedure we will use AIC to select the best ARIMA on train set data. The problem with those techniques is that they are very far from extimate the \"confidence\" of ARIMA predictions. \n",
    "\n",
    "Hence, the cross validation period is used to select the stock/arima pair. How? There are two possibilities: min [MSE](https://en.wikipedia.org/wiki/Mean_squared_error) / max [R2](https://en.wikipedia.org/wiki/Coefficient_of_determination) on cross validation. Minimizing MSE would give more relevance to flat or quasi-flat securities, so we will prefer to use __max R2__. The idea is to BUY the top x stocks with higest R2 cumulative return in cross validation. \n",
    "\n",
    "The test period is for testing those orders in terms of cumulative returns.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "arima_perf = DataFrame({'security':  df.columns.tolist(), \n",
    "                        \n",
    "                        'cumulative_return_test': 0, \n",
    "                        'ARIMA_pred_cumulative_return_test': 0, \n",
    "                        'ARIMA_MSE_cumulative_return_test': 0, \n",
    "                        'ARIMA_R2_cumulative_return_test': 0, \n",
    "                        'ARIMA_MSE_price_test': 0 ,\n",
    "                        'ARIMA_R2_price_test': 0 ,\n",
    "                        \n",
    "                        'cumulative_return_xval': 0, \n",
    "                        'ARIMA_pred_cumulative_return_xval': 0, \n",
    "                        'ARIMA_MSE_cumulative_return_xval': 0, \n",
    "                        'ARIMA_R2_cumulative_return_xval': 0, \n",
    "                        'ARIMA_MSE_price_xval': 0 ,\n",
    "                        'ARIMA_R2_price_xval': 0 , \n",
    "                        \n",
    "                        'ARIMA_best_order_0': 0, \n",
    "                        'ARIMA_best_order_1': 0, \n",
    "                        'ARIMA_best_order_2': 0} )\n",
    "arima_perf = arima_perf.set_index(['security'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 A\n"
     ]
    }
   ],
   "source": [
    "p_tr = 0.60\n",
    "p_xval = 0.20\n",
    "p_test = 1-p_tr-p_xval  \n",
    "\n",
    "for i,sec in enumerate(arima_perf.index.tolist()):\n",
    "    if (i % 100 == 0) or (i ==10):\n",
    "        print(i,sec)\n",
    "    \n",
    "    ## data \n",
    "    series = df[sec]\n",
    "    series = series.tolist()\n",
    "    tr_size = int(len(series) * p_tr)\n",
    "    xval_size = int(len(series) * p_xval)\n",
    "    train, txval, test = series[0:tr_size], series[tr_size:(tr_size+xval_size)] , series[(tr_size+xval_size):len(series)]\n",
    "    \n",
    "    ## fit model \n",
    "    best_aic, best_ord , best_model = get_best_arima(train)\n",
    "    \n",
    "    ## predict, assess \n",
    "    predictions = predict_arima(train,(len(txval)+len(test)),best_ord,verbose=False,do_plot=False)\n",
    "    \n",
    "    ## store\n",
    "    arima_perf.loc[sec,'ARIMA_best_order_0'] = best_ord[0]\n",
    "    arima_perf.loc[sec,'ARIMA_best_order_1'] = best_ord[1]\n",
    "    arima_perf.loc[sec,'ARIMA_best_order_2'] = best_ord[2]\n",
    "    \n",
    "    # xval \n",
    "    pred_cumulative_returns_xval = cumulative_returns(pd.Series(predictions[0:xval_size]))\n",
    "    cumulative_returns_xval = cumulative_returns(pd.Series(txval))\n",
    "    \n",
    "    arima_perf.loc[sec,'cumulative_return_xval'] = cumulative_returns_xval[len(cumulative_returns_xval)-1]\n",
    "    arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval'] = pred_cumulative_returns_xval[len(pred_cumulative_returns_xval)-1]\n",
    "    arima_perf.loc[sec,'ARIMA_MSE_cumulative_return_xval'] = mean_squared_error(cumulative_returns_xval, pred_cumulative_returns_xval)\n",
    "    arima_perf.loc[sec,'ARIMA_R2_cumulative_return_xval'] = r2_score(cumulative_returns_xval, pred_cumulative_returns_xval)\n",
    "    arima_perf.loc[sec,'ARIMA_MSE_price_xval'] = mean_squared_error(txval, predictions[0:xval_size])\n",
    "    arima_perf.loc[sec,'ARIMA_R2_price_xval'] = r2_score(txval, predictions[0:xval_size])\n",
    "    \n",
    "    # test \n",
    "    pred_cumulative_returns_test = cumulative_returns(pd.Series(predictions[xval_size:]))\n",
    "    cumulative_returns_test = cumulative_returns(pd.Series(test))\n",
    "    \n",
    "    arima_perf.loc[sec,'cumulative_return_test'] = cumulative_returns_test[len(cumulative_returns_test)-1]\n",
    "    arima_perf.loc[sec,'ARIMA_pred_cumulative_return_test'] = pred_cumulative_returns_test[len(pred_cumulative_returns_test)-1]\n",
    "    arima_perf.loc[sec,'ARIMA_MSE_cumulative_return_test'] = mean_squared_error(cumulative_returns_test, pred_cumulative_returns_test)\n",
    "    arima_perf.loc[sec,'ARIMA_R2_cumulative_return_test'] = r2_score(cumulative_returns_test, pred_cumulative_returns_test)\n",
    "    arima_perf.loc[sec,'ARIMA_MSE_price_test'] = mean_squared_error(test, predictions[xval_size:])\n",
    "    arima_perf.loc[sec,'ARIMA_R2_price_test'] = r2_score(test, predictions[xval_size:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ARIMA_MSE_cumulative_return_test</th>\n",
       "      <th>ARIMA_MSE_cumulative_return_xval</th>\n",
       "      <th>ARIMA_MSE_price_test</th>\n",
       "      <th>ARIMA_MSE_price_xval</th>\n",
       "      <th>ARIMA_R2_cumulative_return_test</th>\n",
       "      <th>ARIMA_R2_cumulative_return_xval</th>\n",
       "      <th>ARIMA_R2_price_test</th>\n",
       "      <th>ARIMA_R2_price_xval</th>\n",
       "      <th>ARIMA_best_order_0</th>\n",
       "      <th>ARIMA_best_order_1</th>\n",
       "      <th>ARIMA_best_order_2</th>\n",
       "      <th>ARIMA_pred_cumulative_return_test</th>\n",
       "      <th>ARIMA_pred_cumulative_return_xval</th>\n",
       "      <th>cumulative_return_test</th>\n",
       "      <th>cumulative_return_xval</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>security</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AAP</th>\n",
       "      <td>0.001251</td>\n",
       "      <td>0.002538</td>\n",
       "      <td>393.883989</td>\n",
       "      <td>93.583475</td>\n",
       "      <td>-15.170917</td>\n",
       "      <td>-14.126773</td>\n",
       "      <td>-248.649086</td>\n",
       "      <td>-25.287753</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>9.086896e-02</td>\n",
       "      <td>5.227870e-02</td>\n",
       "      <td>0.021916</td>\n",
       "      <td>-0.021418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABB</th>\n",
       "      <td>0.024093</td>\n",
       "      <td>0.010053</td>\n",
       "      <td>108.563000</td>\n",
       "      <td>8.667667</td>\n",
       "      <td>-1185.537134</td>\n",
       "      <td>-182.174342</td>\n",
       "      <td>-8826.681824</td>\n",
       "      <td>-267.787245</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2.679850e-01</td>\n",
       "      <td>1.929188e-01</td>\n",
       "      <td>0.002438</td>\n",
       "      <td>0.018977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AAPL</th>\n",
       "      <td>0.001220</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>65.973485</td>\n",
       "      <td>4.156296</td>\n",
       "      <td>-1.616267</td>\n",
       "      <td>-0.101245</td>\n",
       "      <td>-5.596613</td>\n",
       "      <td>-0.791505</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.714683e-04</td>\n",
       "      <td>-3.094419e-04</td>\n",
       "      <td>0.051173</td>\n",
       "      <td>0.020619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A</th>\n",
       "      <td>0.000048</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>11.752400</td>\n",
       "      <td>5.356483</td>\n",
       "      <td>-1.340708</td>\n",
       "      <td>-1.255402</td>\n",
       "      <td>-179.184130</td>\n",
       "      <td>-26.078473</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.087453e-11</td>\n",
       "      <td>-4.497402e-12</td>\n",
       "      <td>0.001246</td>\n",
       "      <td>0.020434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AAL</th>\n",
       "      <td>0.001021</td>\n",
       "      <td>0.005708</td>\n",
       "      <td>10.718462</td>\n",
       "      <td>12.656813</td>\n",
       "      <td>-0.050973</td>\n",
       "      <td>-3.186094</td>\n",
       "      <td>-4.717228</td>\n",
       "      <td>-3.272436</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.670376e-02</td>\n",
       "      <td>2.286368e-02</td>\n",
       "      <td>0.059654</td>\n",
       "      <td>-0.058786</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          ARIMA_MSE_cumulative_return_test  ARIMA_MSE_cumulative_return_xval  \\\n",
       "security                                                                       \n",
       "AAP                               0.001251                          0.002538   \n",
       "ABB                               0.024093                          0.010053   \n",
       "AAPL                              0.001220                          0.000123   \n",
       "A                                 0.000048                          0.000149   \n",
       "AAL                               0.001021                          0.005708   \n",
       "\n",
       "          ARIMA_MSE_price_test  ARIMA_MSE_price_xval  \\\n",
       "security                                               \n",
       "AAP                 393.883989             93.583475   \n",
       "ABB                 108.563000              8.667667   \n",
       "AAPL                 65.973485              4.156296   \n",
       "A                    11.752400              5.356483   \n",
       "AAL                  10.718462             12.656813   \n",
       "\n",
       "          ARIMA_R2_cumulative_return_test  ARIMA_R2_cumulative_return_xval  \\\n",
       "security                                                                     \n",
       "AAP                            -15.170917                       -14.126773   \n",
       "ABB                          -1185.537134                      -182.174342   \n",
       "AAPL                            -1.616267                        -0.101245   \n",
       "A                               -1.340708                        -1.255402   \n",
       "AAL                             -0.050973                        -3.186094   \n",
       "\n",
       "          ARIMA_R2_price_test  ARIMA_R2_price_xval  ARIMA_best_order_0  \\\n",
       "security                                                                 \n",
       "AAP               -248.649086           -25.287753                   0   \n",
       "ABB              -8826.681824          -267.787245                   4   \n",
       "AAPL                -5.596613            -0.791505                   0   \n",
       "A                 -179.184130           -26.078473                   0   \n",
       "AAL                 -4.717228            -3.272436                   0   \n",
       "\n",
       "          ARIMA_best_order_1  ARIMA_best_order_2  \\\n",
       "security                                           \n",
       "AAP                        2                   1   \n",
       "ABB                        2                   0   \n",
       "AAPL                       1                   0   \n",
       "A                          0                   0   \n",
       "AAL                        1                   0   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_test  \\\n",
       "security                                      \n",
       "AAP                            9.086896e-02   \n",
       "ABB                            2.679850e-01   \n",
       "AAPL                          -3.714683e-04   \n",
       "A                             -3.087453e-11   \n",
       "AAL                            2.670376e-02   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_xval  cumulative_return_test  \\\n",
       "security                                                              \n",
       "AAP                            5.227870e-02                0.021916   \n",
       "ABB                            1.929188e-01                0.002438   \n",
       "AAPL                          -3.094419e-04                0.051173   \n",
       "A                             -4.497402e-12                0.001246   \n",
       "AAL                            2.286368e-02                0.059654   \n",
       "\n",
       "          cumulative_return_xval  \n",
       "security                          \n",
       "AAP                    -0.021418  \n",
       "ABB                     0.018977  \n",
       "AAPL                    0.020619  \n",
       "A                       0.020434  \n",
       "AAL                    -0.058786  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_perf.sort_values(by='ARIMA_MSE_price_test' , ascending=False).head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#arima_perf.to_csv('data/arima_perf.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reloading full results "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "arima_perf = pd.read_csv('data/arima_perf.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stock picking \n",
    "\n",
    "Let's consider securities with R2 cumulative return higher than 65% in cross validation and \n",
    "* __BUY__ if ARIMA_pred_cumulative_return_xval > 0 and 0.7 =< ARIMA_pred_cumulative_return_xval/cumulative_return_xval <= 2 \n",
    "* __SELL__ if ARIMA_pred_cumulative_return_xval < 0 and 0.7 =< ARIMA_pred_cumulative_return_xval/cumulative_return_xval <= 2\n",
    "* do nothing, otherwise  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ARIMA_MSE_cumulative_return_test</th>\n",
       "      <th>ARIMA_MSE_cumulative_return_xval</th>\n",
       "      <th>ARIMA_MSE_price_test</th>\n",
       "      <th>ARIMA_MSE_price_xval</th>\n",
       "      <th>ARIMA_R2_cumulative_return_test</th>\n",
       "      <th>ARIMA_R2_cumulative_return_xval</th>\n",
       "      <th>ARIMA_R2_price_test</th>\n",
       "      <th>ARIMA_R2_price_xval</th>\n",
       "      <th>ARIMA_best_order_0</th>\n",
       "      <th>ARIMA_best_order_1</th>\n",
       "      <th>ARIMA_best_order_2</th>\n",
       "      <th>ARIMA_pred_cumulative_return_test</th>\n",
       "      <th>ARIMA_pred_cumulative_return_xval</th>\n",
       "      <th>cumulative_return_test</th>\n",
       "      <th>cumulative_return_xval</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>security</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>FB</th>\n",
       "      <td>0.001949</td>\n",
       "      <td>0.000030</td>\n",
       "      <td>71.956888</td>\n",
       "      <td>0.646026</td>\n",
       "      <td>-140.437704</td>\n",
       "      <td>0.906867</td>\n",
       "      <td>-225.582970</td>\n",
       "      <td>0.905298</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.061472</td>\n",
       "      <td>0.041966</td>\n",
       "      <td>-0.011594</td>\n",
       "      <td>0.042938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HPQ</th>\n",
       "      <td>0.000063</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.039945</td>\n",
       "      <td>0.001443</td>\n",
       "      <td>0.582156</td>\n",
       "      <td>0.905489</td>\n",
       "      <td>0.250655</td>\n",
       "      <td>0.903739</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.022412</td>\n",
       "      <td>0.019104</td>\n",
       "      <td>0.023923</td>\n",
       "      <td>0.015583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PTR</th>\n",
       "      <td>0.000155</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>0.935566</td>\n",
       "      <td>0.081298</td>\n",
       "      <td>-1.218640</td>\n",
       "      <td>0.842538</td>\n",
       "      <td>-1.779597</td>\n",
       "      <td>0.713089</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.017562</td>\n",
       "      <td>-0.016152</td>\n",
       "      <td>-0.020464</td>\n",
       "      <td>-0.021360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ANSS</th>\n",
       "      <td>0.005583</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>69.935138</td>\n",
       "      <td>0.029596</td>\n",
       "      <td>-4.895083</td>\n",
       "      <td>0.823553</td>\n",
       "      <td>-4.993023</td>\n",
       "      <td>0.834869</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.010247</td>\n",
       "      <td>0.008628</td>\n",
       "      <td>0.080090</td>\n",
       "      <td>0.009839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TFX</th>\n",
       "      <td>0.007363</td>\n",
       "      <td>0.000055</td>\n",
       "      <td>402.987962</td>\n",
       "      <td>0.981110</td>\n",
       "      <td>-10.095887</td>\n",
       "      <td>0.784627</td>\n",
       "      <td>-12.839906</td>\n",
       "      <td>0.906103</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0.048541</td>\n",
       "      <td>0.037775</td>\n",
       "      <td>-0.074426</td>\n",
       "      <td>0.042468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TROW</th>\n",
       "      <td>0.000281</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>10.991321</td>\n",
       "      <td>6.968545</td>\n",
       "      <td>-11.416958</td>\n",
       "      <td>0.752470</td>\n",
       "      <td>-93.103668</td>\n",
       "      <td>-21.231373</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0.020896</td>\n",
       "      <td>0.020089</td>\n",
       "      <td>-0.008489</td>\n",
       "      <td>0.025722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CVE</th>\n",
       "      <td>0.002503</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.391605</td>\n",
       "      <td>0.014616</td>\n",
       "      <td>-5.280363</td>\n",
       "      <td>0.750478</td>\n",
       "      <td>-9.441100</td>\n",
       "      <td>0.786076</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.088220</td>\n",
       "      <td>-0.067557</td>\n",
       "      <td>-0.017526</td>\n",
       "      <td>-0.072312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSM</th>\n",
       "      <td>0.000413</td>\n",
       "      <td>0.000053</td>\n",
       "      <td>11.665200</td>\n",
       "      <td>1.876501</td>\n",
       "      <td>-12.393980</td>\n",
       "      <td>0.737158</td>\n",
       "      <td>-46.311200</td>\n",
       "      <td>-0.096968</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.033369</td>\n",
       "      <td>-0.026909</td>\n",
       "      <td>-0.009389</td>\n",
       "      <td>-0.028382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SNN</th>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.125051</td>\n",
       "      <td>0.052441</td>\n",
       "      <td>-0.093462</td>\n",
       "      <td>0.730933</td>\n",
       "      <td>0.167980</td>\n",
       "      <td>-0.229658</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.015842</td>\n",
       "      <td>0.013414</td>\n",
       "      <td>0.025927</td>\n",
       "      <td>0.019781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TTWO</th>\n",
       "      <td>0.000880</td>\n",
       "      <td>0.000024</td>\n",
       "      <td>2.810949</td>\n",
       "      <td>0.238946</td>\n",
       "      <td>-0.004700</td>\n",
       "      <td>0.722089</td>\n",
       "      <td>0.216105</td>\n",
       "      <td>0.288050</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.020555</td>\n",
       "      <td>0.017479</td>\n",
       "      <td>0.071886</td>\n",
       "      <td>0.023311</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          ARIMA_MSE_cumulative_return_test  ARIMA_MSE_cumulative_return_xval  \\\n",
       "security                                                                       \n",
       "FB                                0.001949                          0.000030   \n",
       "HPQ                               0.000063                          0.000004   \n",
       "PTR                               0.000155                          0.000009   \n",
       "ANSS                              0.005583                          0.000003   \n",
       "TFX                               0.007363                          0.000055   \n",
       "TROW                              0.000281                          0.000016   \n",
       "CVE                               0.002503                          0.000154   \n",
       "MSM                               0.000413                          0.000053   \n",
       "SNN                               0.000149                          0.000011   \n",
       "TTWO                              0.000880                          0.000024   \n",
       "\n",
       "          ARIMA_MSE_price_test  ARIMA_MSE_price_xval  \\\n",
       "security                                               \n",
       "FB                   71.956888              0.646026   \n",
       "HPQ                   0.039945              0.001443   \n",
       "PTR                   0.935566              0.081298   \n",
       "ANSS                 69.935138              0.029596   \n",
       "TFX                 402.987962              0.981110   \n",
       "TROW                 10.991321              6.968545   \n",
       "CVE                   0.391605              0.014616   \n",
       "MSM                  11.665200              1.876501   \n",
       "SNN                   0.125051              0.052441   \n",
       "TTWO                  2.810949              0.238946   \n",
       "\n",
       "          ARIMA_R2_cumulative_return_test  ARIMA_R2_cumulative_return_xval  \\\n",
       "security                                                                     \n",
       "FB                            -140.437704                         0.906867   \n",
       "HPQ                              0.582156                         0.905489   \n",
       "PTR                             -1.218640                         0.842538   \n",
       "ANSS                            -4.895083                         0.823553   \n",
       "TFX                            -10.095887                         0.784627   \n",
       "TROW                           -11.416958                         0.752470   \n",
       "CVE                             -5.280363                         0.750478   \n",
       "MSM                            -12.393980                         0.737158   \n",
       "SNN                             -0.093462                         0.730933   \n",
       "TTWO                            -0.004700                         0.722089   \n",
       "\n",
       "          ARIMA_R2_price_test  ARIMA_R2_price_xval  ARIMA_best_order_0  \\\n",
       "security                                                                 \n",
       "FB                -225.582970             0.905298                   2   \n",
       "HPQ                  0.250655             0.903739                   0   \n",
       "PTR                 -1.779597             0.713089                   0   \n",
       "ANSS                -4.993023             0.834869                   0   \n",
       "TFX                -12.839906             0.906103                   2   \n",
       "TROW               -93.103668           -21.231373                   0   \n",
       "CVE                 -9.441100             0.786076                   0   \n",
       "MSM                -46.311200            -0.096968                   0   \n",
       "SNN                  0.167980            -0.229658                   0   \n",
       "TTWO                 0.216105             0.288050                   0   \n",
       "\n",
       "          ARIMA_best_order_1  ARIMA_best_order_2  \\\n",
       "security                                           \n",
       "FB                         2                   0   \n",
       "HPQ                        1                   0   \n",
       "PTR                        1                   2   \n",
       "ANSS                       1                   0   \n",
       "TFX                        2                   2   \n",
       "TROW                       1                   3   \n",
       "CVE                        1                   0   \n",
       "MSM                        1                   0   \n",
       "SNN                        1                   0   \n",
       "TTWO                       1                   1   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_test  \\\n",
       "security                                      \n",
       "FB                                 0.061472   \n",
       "HPQ                                0.022412   \n",
       "PTR                               -0.017562   \n",
       "ANSS                               0.010247   \n",
       "TFX                                0.048541   \n",
       "TROW                               0.020896   \n",
       "CVE                               -0.088220   \n",
       "MSM                               -0.033369   \n",
       "SNN                                0.015842   \n",
       "TTWO                               0.020555   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_xval  cumulative_return_test  \\\n",
       "security                                                              \n",
       "FB                                 0.041966               -0.011594   \n",
       "HPQ                                0.019104                0.023923   \n",
       "PTR                               -0.016152               -0.020464   \n",
       "ANSS                               0.008628                0.080090   \n",
       "TFX                                0.037775               -0.074426   \n",
       "TROW                               0.020089               -0.008489   \n",
       "CVE                               -0.067557               -0.017526   \n",
       "MSM                               -0.026909               -0.009389   \n",
       "SNN                                0.013414                0.025927   \n",
       "TTWO                               0.017479                0.071886   \n",
       "\n",
       "          cumulative_return_xval  \n",
       "security                          \n",
       "FB                      0.042938  \n",
       "HPQ                     0.015583  \n",
       "PTR                    -0.021360  \n",
       "ANSS                    0.009839  \n",
       "TFX                     0.042468  \n",
       "TROW                    0.025722  \n",
       "CVE                    -0.072312  \n",
       "MSM                    -0.028382  \n",
       "SNN                     0.019781  \n",
       "TTWO                    0.023311  "
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_perf = arima_perf.set_index('security')\n",
    "arima_perf.sort_values(by='ARIMA_R2_cumulative_return_xval' , ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "arima_perf['ACTION'] = ''\n",
    "for i,sec in enumerate(arima_perf.index.tolist()):\n",
    "    if arima_perf.loc[sec,'ARIMA_R2_cumulative_return_xval'] > 0.65:\n",
    "        \n",
    "        if arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval'] > 0 and arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval']/arima_perf.loc[sec,'cumulative_return_xval'] >= 0.7 and arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval']/arima_perf.loc[sec,'cumulative_return_xval'] <= 2:\n",
    "            arima_perf.loc[sec,'ACTION'] = 'BUY'\n",
    "            \n",
    "        if arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval'] < 0 and arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval']/arima_perf.loc[sec,'cumulative_return_xval'] >= 0.7 and arima_perf.loc[sec,'ARIMA_pred_cumulative_return_xval']/arima_perf.loc[sec,'cumulative_return_xval'] <= 2:\n",
    "            arima_perf.loc[sec,'ACTION'] = 'SELL'    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ACTION</th>\n",
       "      <th>ARIMA_R2_cumulative_return_xval</th>\n",
       "      <th>ARIMA_pred_cumulative_return_xval</th>\n",
       "      <th>cumulative_return_xval</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>security</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>FB</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.906867</td>\n",
       "      <td>0.041966</td>\n",
       "      <td>0.042938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HPQ</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.905489</td>\n",
       "      <td>0.019104</td>\n",
       "      <td>0.015583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PTR</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.842538</td>\n",
       "      <td>-0.016152</td>\n",
       "      <td>-0.021360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ANSS</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.823553</td>\n",
       "      <td>0.008628</td>\n",
       "      <td>0.009839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TFX</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.784627</td>\n",
       "      <td>0.037775</td>\n",
       "      <td>0.042468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TROW</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.752470</td>\n",
       "      <td>0.020089</td>\n",
       "      <td>0.025722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CVE</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.750478</td>\n",
       "      <td>-0.067557</td>\n",
       "      <td>-0.072312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSM</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.737158</td>\n",
       "      <td>-0.026909</td>\n",
       "      <td>-0.028382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SNN</th>\n",
       "      <td></td>\n",
       "      <td>0.730933</td>\n",
       "      <td>0.013414</td>\n",
       "      <td>0.019781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TTWO</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.722089</td>\n",
       "      <td>0.017479</td>\n",
       "      <td>0.023311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSFT</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.720882</td>\n",
       "      <td>0.019285</td>\n",
       "      <td>0.020318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLR</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.710080</td>\n",
       "      <td>-0.093355</td>\n",
       "      <td>-0.071057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CNK</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.703598</td>\n",
       "      <td>-0.025912</td>\n",
       "      <td>-0.031714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GIL</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.689751</td>\n",
       "      <td>0.014245</td>\n",
       "      <td>0.012926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SJM</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.687605</td>\n",
       "      <td>-0.013097</td>\n",
       "      <td>-0.017805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RSPP</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.682839</td>\n",
       "      <td>-0.022099</td>\n",
       "      <td>-0.028948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCLN</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.676259</td>\n",
       "      <td>0.032062</td>\n",
       "      <td>0.045262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RPM</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.673798</td>\n",
       "      <td>-0.009319</td>\n",
       "      <td>-0.013198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CBSH</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.653150</td>\n",
       "      <td>-0.007920</td>\n",
       "      <td>-0.010728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ORAN</th>\n",
       "      <td>SELL</td>\n",
       "      <td>0.652198</td>\n",
       "      <td>-0.021244</td>\n",
       "      <td>-0.020000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ACTION  ARIMA_R2_cumulative_return_xval  \\\n",
       "security                                           \n",
       "FB          BUY                         0.906867   \n",
       "HPQ         BUY                         0.905489   \n",
       "PTR        SELL                         0.842538   \n",
       "ANSS        BUY                         0.823553   \n",
       "TFX         BUY                         0.784627   \n",
       "TROW        BUY                         0.752470   \n",
       "CVE        SELL                         0.750478   \n",
       "MSM        SELL                         0.737158   \n",
       "SNN                                     0.730933   \n",
       "TTWO        BUY                         0.722089   \n",
       "MSFT        BUY                         0.720882   \n",
       "CLR        SELL                         0.710080   \n",
       "CNK        SELL                         0.703598   \n",
       "GIL         BUY                         0.689751   \n",
       "SJM        SELL                         0.687605   \n",
       "RSPP       SELL                         0.682839   \n",
       "PCLN        BUY                         0.676259   \n",
       "RPM        SELL                         0.673798   \n",
       "CBSH       SELL                         0.653150   \n",
       "ORAN       SELL                         0.652198   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_xval  cumulative_return_xval  \n",
       "security                                                             \n",
       "FB                                 0.041966                0.042938  \n",
       "HPQ                                0.019104                0.015583  \n",
       "PTR                               -0.016152               -0.021360  \n",
       "ANSS                               0.008628                0.009839  \n",
       "TFX                                0.037775                0.042468  \n",
       "TROW                               0.020089                0.025722  \n",
       "CVE                               -0.067557               -0.072312  \n",
       "MSM                               -0.026909               -0.028382  \n",
       "SNN                                0.013414                0.019781  \n",
       "TTWO                               0.017479                0.023311  \n",
       "MSFT                               0.019285                0.020318  \n",
       "CLR                               -0.093355               -0.071057  \n",
       "CNK                               -0.025912               -0.031714  \n",
       "GIL                                0.014245                0.012926  \n",
       "SJM                               -0.013097               -0.017805  \n",
       "RSPP                              -0.022099               -0.028948  \n",
       "PCLN                               0.032062                0.045262  \n",
       "RPM                               -0.009319               -0.013198  \n",
       "CBSH                              -0.007920               -0.010728  \n",
       "ORAN                              -0.021244               -0.020000  "
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_perf = arima_perf.sort_values(by='ARIMA_R2_cumulative_return_xval' , ascending=False)\n",
    "arima_perf[['ACTION','ARIMA_R2_cumulative_return_xval','ARIMA_pred_cumulative_return_xval','cumulative_return_xval']][:20]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice that we should not buy SNN (Smith & Nephew PLC) because the predicted cumulative return is too small vs. the actual cumulative return. Is this reasonable? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Note #1: taking just predicted cumulative return would lead to losses "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ACTION</th>\n",
       "      <th>ARIMA_R2_cumulative_return_xval</th>\n",
       "      <th>ARIMA_pred_cumulative_return_xval</th>\n",
       "      <th>cumulative_return_xval</th>\n",
       "      <th>UNEXPECTED_LOSS</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>security</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DB</th>\n",
       "      <td></td>\n",
       "      <td>-266.040093</td>\n",
       "      <td>0.366728</td>\n",
       "      <td>-0.011898</td>\n",
       "      <td>-0.378626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HAS</th>\n",
       "      <td></td>\n",
       "      <td>-659.516643</td>\n",
       "      <td>0.322470</td>\n",
       "      <td>-0.013139</td>\n",
       "      <td>-0.335610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CRH</th>\n",
       "      <td></td>\n",
       "      <td>-291.629287</td>\n",
       "      <td>0.260132</td>\n",
       "      <td>0.026519</td>\n",
       "      <td>-0.233613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BCR</th>\n",
       "      <td></td>\n",
       "      <td>-490.134359</td>\n",
       "      <td>0.222576</td>\n",
       "      <td>0.011541</td>\n",
       "      <td>-0.211035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABB</th>\n",
       "      <td></td>\n",
       "      <td>-182.174354</td>\n",
       "      <td>0.192919</td>\n",
       "      <td>0.018977</td>\n",
       "      <td>-0.173942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XLNX</th>\n",
       "      <td></td>\n",
       "      <td>-8.568886</td>\n",
       "      <td>0.178509</td>\n",
       "      <td>0.032954</td>\n",
       "      <td>-0.145555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LULU</th>\n",
       "      <td></td>\n",
       "      <td>-114.901407</td>\n",
       "      <td>0.151673</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.151673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MYL</th>\n",
       "      <td></td>\n",
       "      <td>-35.941912</td>\n",
       "      <td>0.145565</td>\n",
       "      <td>0.035821</td>\n",
       "      <td>-0.109744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HON</th>\n",
       "      <td></td>\n",
       "      <td>-139.994993</td>\n",
       "      <td>0.125998</td>\n",
       "      <td>0.014617</td>\n",
       "      <td>-0.111381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SNA</th>\n",
       "      <td></td>\n",
       "      <td>-49.570311</td>\n",
       "      <td>0.118654</td>\n",
       "      <td>-0.034591</td>\n",
       "      <td>-0.153246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SHOP</th>\n",
       "      <td></td>\n",
       "      <td>-1.231333</td>\n",
       "      <td>0.113252</td>\n",
       "      <td>0.082583</td>\n",
       "      <td>-0.030669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LRCX</th>\n",
       "      <td></td>\n",
       "      <td>-18.379027</td>\n",
       "      <td>0.109453</td>\n",
       "      <td>0.024424</td>\n",
       "      <td>-0.085030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>STI</th>\n",
       "      <td></td>\n",
       "      <td>-104.425721</td>\n",
       "      <td>0.107933</td>\n",
       "      <td>-0.000699</td>\n",
       "      <td>-0.108633</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALK</th>\n",
       "      <td></td>\n",
       "      <td>-18.278192</td>\n",
       "      <td>0.106975</td>\n",
       "      <td>-0.053927</td>\n",
       "      <td>-0.160901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ORLY</th>\n",
       "      <td></td>\n",
       "      <td>-9.785214</td>\n",
       "      <td>0.105915</td>\n",
       "      <td>-0.068722</td>\n",
       "      <td>-0.174638</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OLN</th>\n",
       "      <td></td>\n",
       "      <td>-26.820509</td>\n",
       "      <td>0.105271</td>\n",
       "      <td>-0.048002</td>\n",
       "      <td>-0.153274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AKAM</th>\n",
       "      <td></td>\n",
       "      <td>-18.052599</td>\n",
       "      <td>0.104699</td>\n",
       "      <td>0.011327</td>\n",
       "      <td>-0.093372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PHG</th>\n",
       "      <td></td>\n",
       "      <td>-95.027062</td>\n",
       "      <td>0.102063</td>\n",
       "      <td>0.008360</td>\n",
       "      <td>-0.093703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PX</th>\n",
       "      <td></td>\n",
       "      <td>-533.086830</td>\n",
       "      <td>0.100760</td>\n",
       "      <td>0.006977</td>\n",
       "      <td>-0.093783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLT</th>\n",
       "      <td></td>\n",
       "      <td>-7.524866</td>\n",
       "      <td>0.097465</td>\n",
       "      <td>-0.072954</td>\n",
       "      <td>-0.170419</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ACTION  ARIMA_R2_cumulative_return_xval  \\\n",
       "security                                           \n",
       "DB                                   -266.040093   \n",
       "HAS                                  -659.516643   \n",
       "CRH                                  -291.629287   \n",
       "BCR                                  -490.134359   \n",
       "ABB                                  -182.174354   \n",
       "XLNX                                   -8.568886   \n",
       "LULU                                 -114.901407   \n",
       "MYL                                   -35.941912   \n",
       "HON                                  -139.994993   \n",
       "SNA                                   -49.570311   \n",
       "SHOP                                   -1.231333   \n",
       "LRCX                                  -18.379027   \n",
       "STI                                  -104.425721   \n",
       "ALK                                   -18.278192   \n",
       "ORLY                                   -9.785214   \n",
       "OLN                                   -26.820509   \n",
       "AKAM                                  -18.052599   \n",
       "PHG                                   -95.027062   \n",
       "PX                                   -533.086830   \n",
       "FLT                                    -7.524866   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_xval  cumulative_return_xval  \\\n",
       "security                                                              \n",
       "DB                                 0.366728               -0.011898   \n",
       "HAS                                0.322470               -0.013139   \n",
       "CRH                                0.260132                0.026519   \n",
       "BCR                                0.222576                0.011541   \n",
       "ABB                                0.192919                0.018977   \n",
       "XLNX                               0.178509                0.032954   \n",
       "LULU                               0.151673                0.000000   \n",
       "MYL                                0.145565                0.035821   \n",
       "HON                                0.125998                0.014617   \n",
       "SNA                                0.118654               -0.034591   \n",
       "SHOP                               0.113252                0.082583   \n",
       "LRCX                               0.109453                0.024424   \n",
       "STI                                0.107933               -0.000699   \n",
       "ALK                                0.106975               -0.053927   \n",
       "ORLY                               0.105915               -0.068722   \n",
       "OLN                                0.105271               -0.048002   \n",
       "AKAM                               0.104699                0.011327   \n",
       "PHG                                0.102063                0.008360   \n",
       "PX                                 0.100760                0.006977   \n",
       "FLT                                0.097465               -0.072954   \n",
       "\n",
       "          UNEXPECTED_LOSS  \n",
       "security                   \n",
       "DB              -0.378626  \n",
       "HAS             -0.335610  \n",
       "CRH             -0.233613  \n",
       "BCR             -0.211035  \n",
       "ABB             -0.173942  \n",
       "XLNX            -0.145555  \n",
       "LULU            -0.151673  \n",
       "MYL             -0.109744  \n",
       "HON             -0.111381  \n",
       "SNA             -0.153246  \n",
       "SHOP            -0.030669  \n",
       "LRCX            -0.085030  \n",
       "STI             -0.108633  \n",
       "ALK             -0.160901  \n",
       "ORLY            -0.174638  \n",
       "OLN             -0.153274  \n",
       "AKAM            -0.093372  \n",
       "PHG             -0.093703  \n",
       "PX              -0.093783  \n",
       "FLT             -0.170419  "
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_perf = arima_perf.sort_values(by='ARIMA_pred_cumulative_return_xval' , ascending=False)\n",
    "loss = arima_perf[['ACTION','ARIMA_R2_cumulative_return_xval','ARIMA_pred_cumulative_return_xval','cumulative_return_xval']][:20]\n",
    "loss['UNEXPECTED_LOSS'] = loss['cumulative_return_xval'] - loss['ARIMA_pred_cumulative_return_xval']\n",
    "loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Note #2: using MSE instead of R2 would give relevance to flat or quasi-flat securities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ACTION</th>\n",
       "      <th>ARIMA_R2_cumulative_return_xval</th>\n",
       "      <th>ARIMA_MSE_cumulative_return_xval</th>\n",
       "      <th>ARIMA_pred_cumulative_return_xval</th>\n",
       "      <th>cumulative_return_xval</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>security</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MJN</th>\n",
       "      <td></td>\n",
       "      <td>-11.308938</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>-4.373945e-04</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MBLY</th>\n",
       "      <td></td>\n",
       "      <td>-1.526417</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>2.993046e-04</td>\n",
       "      <td>-0.001454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BK</th>\n",
       "      <td></td>\n",
       "      <td>-0.129741</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>2.466738e-11</td>\n",
       "      <td>0.001702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ARMK</th>\n",
       "      <td></td>\n",
       "      <td>0.528055</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>-4.930150e-03</td>\n",
       "      <td>-0.005441</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ANSS</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.823553</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>8.627615e-03</td>\n",
       "      <td>0.009839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABT</th>\n",
       "      <td></td>\n",
       "      <td>0.390381</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>-3.331715e-03</td>\n",
       "      <td>-0.003656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BRCD</th>\n",
       "      <td></td>\n",
       "      <td>-0.862069</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-0.000795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HPQ</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.905489</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>1.910449e-02</td>\n",
       "      <td>0.015583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VAL</th>\n",
       "      <td></td>\n",
       "      <td>-69.365377</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>3.558071e-03</td>\n",
       "      <td>0.000534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PPL</th>\n",
       "      <td></td>\n",
       "      <td>-0.338807</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>-5.536341e-03</td>\n",
       "      <td>-0.003145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OA</th>\n",
       "      <td></td>\n",
       "      <td>0.082860</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>-4.567830e-03</td>\n",
       "      <td>-0.001716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WOOF</th>\n",
       "      <td></td>\n",
       "      <td>-4.103503</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>-1.058252e-03</td>\n",
       "      <td>0.002949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ARCC</th>\n",
       "      <td></td>\n",
       "      <td>0.380232</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>3.279090e-03</td>\n",
       "      <td>0.004585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CMS</th>\n",
       "      <td></td>\n",
       "      <td>0.178759</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>-2.790256e-03</td>\n",
       "      <td>-0.004175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KO</th>\n",
       "      <td></td>\n",
       "      <td>0.166369</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>5.932629e-03</td>\n",
       "      <td>0.006495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SPY</th>\n",
       "      <td></td>\n",
       "      <td>-6.389177</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>2.801555e-03</td>\n",
       "      <td>0.000922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AON</th>\n",
       "      <td></td>\n",
       "      <td>-1.090449</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>-1.544875e-12</td>\n",
       "      <td>0.001754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCK</th>\n",
       "      <td></td>\n",
       "      <td>0.375824</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>6.559811e-03</td>\n",
       "      <td>0.008226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XEL</th>\n",
       "      <td></td>\n",
       "      <td>-0.056521</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>-2.930949e-03</td>\n",
       "      <td>-0.004443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FDS</th>\n",
       "      <td></td>\n",
       "      <td>-1.991265</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>-2.453239e-03</td>\n",
       "      <td>0.000736</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ACTION  ARIMA_R2_cumulative_return_xval  \\\n",
       "security                                           \n",
       "MJN                                   -11.308938   \n",
       "MBLY                                   -1.526417   \n",
       "BK                                     -0.129741   \n",
       "ARMK                                    0.528055   \n",
       "ANSS        BUY                         0.823553   \n",
       "ABT                                     0.390381   \n",
       "BRCD                                   -0.862069   \n",
       "HPQ         BUY                         0.905489   \n",
       "VAL                                   -69.365377   \n",
       "PPL                                    -0.338807   \n",
       "OA                                      0.082860   \n",
       "WOOF                                   -4.103503   \n",
       "ARCC                                    0.380232   \n",
       "CMS                                     0.178759   \n",
       "KO                                      0.166369   \n",
       "SPY                                    -6.389177   \n",
       "AON                                    -1.090449   \n",
       "CCK                                     0.375824   \n",
       "XEL                                    -0.056521   \n",
       "FDS                                    -1.991265   \n",
       "\n",
       "          ARIMA_MSE_cumulative_return_xval  ARIMA_pred_cumulative_return_xval  \\\n",
       "security                                                                        \n",
       "MJN                               0.000001                      -4.373945e-04   \n",
       "MBLY                              0.000002                       2.993046e-04   \n",
       "BK                                0.000002                       2.466738e-11   \n",
       "ARMK                              0.000003                      -4.930150e-03   \n",
       "ANSS                              0.000003                       8.627615e-03   \n",
       "ABT                               0.000003                      -3.331715e-03   \n",
       "BRCD                              0.000004                       0.000000e+00   \n",
       "HPQ                               0.000004                       1.910449e-02   \n",
       "VAL                               0.000004                       3.558071e-03   \n",
       "PPL                               0.000005                      -5.536341e-03   \n",
       "OA                                0.000005                      -4.567830e-03   \n",
       "WOOF                              0.000005                      -1.058252e-03   \n",
       "ARCC                              0.000006                       3.279090e-03   \n",
       "CMS                               0.000006                      -2.790256e-03   \n",
       "KO                                0.000006                       5.932629e-03   \n",
       "SPY                               0.000007                       2.801555e-03   \n",
       "AON                               0.000007                      -1.544875e-12   \n",
       "CCK                               0.000007                       6.559811e-03   \n",
       "XEL                               0.000009                      -2.930949e-03   \n",
       "FDS                               0.000009                      -2.453239e-03   \n",
       "\n",
       "          cumulative_return_xval  \n",
       "security                          \n",
       "MJN                     0.000000  \n",
       "MBLY                   -0.001454  \n",
       "BK                      0.001702  \n",
       "ARMK                   -0.005441  \n",
       "ANSS                    0.009839  \n",
       "ABT                    -0.003656  \n",
       "BRCD                   -0.000795  \n",
       "HPQ                     0.015583  \n",
       "VAL                     0.000534  \n",
       "PPL                    -0.003145  \n",
       "OA                     -0.001716  \n",
       "WOOF                    0.002949  \n",
       "ARCC                    0.004585  \n",
       "CMS                    -0.004175  \n",
       "KO                      0.006495  \n",
       "SPY                     0.000922  \n",
       "AON                     0.001754  \n",
       "CCK                     0.008226  \n",
       "XEL                    -0.004443  \n",
       "FDS                     0.000736  "
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_perf = arima_perf.sort_values(by='ARIMA_MSE_cumulative_return_xval' , ascending=True)\n",
    "flat = arima_perf[['ACTION','ARIMA_R2_cumulative_return_xval','ARIMA_MSE_cumulative_return_xval','ARIMA_pred_cumulative_return_xval','cumulative_return_xval']][:20]\n",
    "flat"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gain/Loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ACTION</th>\n",
       "      <th>ARIMA_R2_cumulative_return_test</th>\n",
       "      <th>ARIMA_pred_cumulative_return_test</th>\n",
       "      <th>cumulative_return_test</th>\n",
       "      <th>EX_POST</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>security</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>FB</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-140.437704</td>\n",
       "      <td>0.061472</td>\n",
       "      <td>-0.011594</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HPQ</th>\n",
       "      <td>BUY</td>\n",
       "      <td>0.582156</td>\n",
       "      <td>0.022412</td>\n",
       "      <td>0.023923</td>\n",
       "      <td>CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PTR</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-1.218640</td>\n",
       "      <td>-0.017562</td>\n",
       "      <td>-0.020464</td>\n",
       "      <td>CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ANSS</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-4.895083</td>\n",
       "      <td>0.010247</td>\n",
       "      <td>0.080090</td>\n",
       "      <td>CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TFX</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-10.095887</td>\n",
       "      <td>0.048541</td>\n",
       "      <td>-0.074426</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TROW</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-11.416958</td>\n",
       "      <td>0.020896</td>\n",
       "      <td>-0.008489</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CVE</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-5.280363</td>\n",
       "      <td>-0.088220</td>\n",
       "      <td>-0.017526</td>\n",
       "      <td>ALMOST CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSM</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-12.393980</td>\n",
       "      <td>-0.033369</td>\n",
       "      <td>-0.009389</td>\n",
       "      <td>ALMOST CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TTWO</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-0.004700</td>\n",
       "      <td>0.020555</td>\n",
       "      <td>0.071886</td>\n",
       "      <td>CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSFT</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-10.025188</td>\n",
       "      <td>0.015215</td>\n",
       "      <td>-0.008975</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLR</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-39.121066</td>\n",
       "      <td>-0.192883</td>\n",
       "      <td>-0.003797</td>\n",
       "      <td>ALMOST CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CNK</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-16.918997</td>\n",
       "      <td>0.023339</td>\n",
       "      <td>-0.013409</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GIL</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-2.566573</td>\n",
       "      <td>0.016807</td>\n",
       "      <td>0.034820</td>\n",
       "      <td>CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SJM</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-23.169909</td>\n",
       "      <td>-0.015968</td>\n",
       "      <td>0.006657</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RSPP</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-2.282108</td>\n",
       "      <td>-0.027242</td>\n",
       "      <td>0.031568</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCLN</th>\n",
       "      <td>BUY</td>\n",
       "      <td>-3.619299</td>\n",
       "      <td>0.055381</td>\n",
       "      <td>-0.033690</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RPM</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-1.869223</td>\n",
       "      <td>-0.011309</td>\n",
       "      <td>-0.005749</td>\n",
       "      <td>ALMOST CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CBSH</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-0.156807</td>\n",
       "      <td>-0.000464</td>\n",
       "      <td>-0.011839</td>\n",
       "      <td>CORRECT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ORAN</th>\n",
       "      <td>SELL</td>\n",
       "      <td>-4.444562</td>\n",
       "      <td>-0.001640</td>\n",
       "      <td>0.028115</td>\n",
       "      <td>WRONG</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ACTION  ARIMA_R2_cumulative_return_test  \\\n",
       "security                                           \n",
       "FB          BUY                      -140.437704   \n",
       "HPQ         BUY                         0.582156   \n",
       "PTR        SELL                        -1.218640   \n",
       "ANSS        BUY                        -4.895083   \n",
       "TFX         BUY                       -10.095887   \n",
       "TROW        BUY                       -11.416958   \n",
       "CVE        SELL                        -5.280363   \n",
       "MSM        SELL                       -12.393980   \n",
       "TTWO        BUY                        -0.004700   \n",
       "MSFT        BUY                       -10.025188   \n",
       "CLR        SELL                       -39.121066   \n",
       "CNK        SELL                       -16.918997   \n",
       "GIL         BUY                        -2.566573   \n",
       "SJM        SELL                       -23.169909   \n",
       "RSPP       SELL                        -2.282108   \n",
       "PCLN        BUY                        -3.619299   \n",
       "RPM        SELL                        -1.869223   \n",
       "CBSH       SELL                        -0.156807   \n",
       "ORAN       SELL                        -4.444562   \n",
       "\n",
       "          ARIMA_pred_cumulative_return_test  cumulative_return_test  \\\n",
       "security                                                              \n",
       "FB                                 0.061472               -0.011594   \n",
       "HPQ                                0.022412                0.023923   \n",
       "PTR                               -0.017562               -0.020464   \n",
       "ANSS                               0.010247                0.080090   \n",
       "TFX                                0.048541               -0.074426   \n",
       "TROW                               0.020896               -0.008489   \n",
       "CVE                               -0.088220               -0.017526   \n",
       "MSM                               -0.033369               -0.009389   \n",
       "TTWO                               0.020555                0.071886   \n",
       "MSFT                               0.015215               -0.008975   \n",
       "CLR                               -0.192883               -0.003797   \n",
       "CNK                                0.023339               -0.013409   \n",
       "GIL                                0.016807                0.034820   \n",
       "SJM                               -0.015968                0.006657   \n",
       "RSPP                              -0.027242                0.031568   \n",
       "PCLN                               0.055381               -0.033690   \n",
       "RPM                               -0.011309               -0.005749   \n",
       "CBSH                              -0.000464               -0.011839   \n",
       "ORAN                              -0.001640                0.028115   \n",
       "\n",
       "                 EX_POST  \n",
       "security                  \n",
       "FB                 WRONG  \n",
       "HPQ              CORRECT  \n",
       "PTR              CORRECT  \n",
       "ANSS             CORRECT  \n",
       "TFX                WRONG  \n",
       "TROW               WRONG  \n",
       "CVE       ALMOST CORRECT  \n",
       "MSM       ALMOST CORRECT  \n",
       "TTWO             CORRECT  \n",
       "MSFT               WRONG  \n",
       "CLR       ALMOST CORRECT  \n",
       "CNK                WRONG  \n",
       "GIL              CORRECT  \n",
       "SJM                WRONG  \n",
       "RSPP               WRONG  \n",
       "PCLN               WRONG  \n",
       "RPM       ALMOST CORRECT  \n",
       "CBSH             CORRECT  \n",
       "ORAN               WRONG  "
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_perf = arima_perf.sort_values(by='ARIMA_R2_cumulative_return_xval' , ascending=False)\n",
    "gain_loss = arima_perf[arima_perf['ACTION'] != ''][['ACTION','ARIMA_R2_cumulative_return_test','ARIMA_pred_cumulative_return_test','cumulative_return_test']]\n",
    "gain_loss['EX_POST'] = ''\n",
    "for i,sec in enumerate(gain_loss.index.tolist()):\n",
    "    if gain_loss.loc[sec,'cumulative_return_test'] * gain_loss.loc[sec,'ARIMA_pred_cumulative_return_test'] < 0:\n",
    "        gain_loss.loc[sec,'EX_POST'] = 'WRONG'\n",
    "    elif np.absolute(gain_loss.loc[sec,'cumulative_return_test']) <  np.absolute(gain_loss.loc[sec,'ARIMA_pred_cumulative_return_test']):\n",
    "        gain_loss.loc[sec,'EX_POST'] = 'ALMOST CORRECT'\n",
    "    else:\n",
    "        gain_loss.loc[sec,'EX_POST'] = 'CORRECT'\n",
    "gain_loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The percentage of mistakes would be: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.47368421052631576"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gain_loss[gain_loss['EX_POST'] == 'WRONG'].shape[0]/gain_loss.shape[0]"
   ]
  }
 ],
 "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": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}