11_decision_trees_random_forests/00_data_prep.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How to transform data into factors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on a conceptual understanding of key factor categories, their rationale and popular metrics, a key task is to identify new factors that may better capture the risks embodied by the return drivers laid out previously, or to find new ones. \n",
    "\n",
    "In either case, it will be important to compare the performance of innovative factors to that of known factors to identify incremental signal gains."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We create the dataset here and store it in our [data](../../data) folder to facilitate reuse in later chapters."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports & Settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-16T00:32:11.108720Z",
     "start_time": "2021-04-16T00:32:11.106680Z"
    }
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-16T00:32:12.061911Z",
     "start_time": "2021-04-16T00:32:11.317486Z"
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pandas_datareader.data as web\n",
    "\n",
    "# from pyfinance.ols import PandasRollingOLS\n",
    "# replaces pyfinance.ols.PandasRollingOLS (no longer maintained)\n",
    "from statsmodels.regression.rolling import RollingOLS\n",
    "import statsmodels.api as sm\n",
    "from talib import RSI, BBANDS, MACD, NATR, ATR\n",
    "\n",
    "from sklearn.feature_selection import mutual_info_classif, mutual_info_regression\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-16T00:32:12.066329Z",
     "start_time": "2021-04-16T00:32:12.063370Z"
    }
   },
   "outputs": [],
   "source": [
    "sns.set_style('whitegrid')\n",
    "idx = pd.IndexSlice"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load US equity OHLCV data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `assets.h5` store can be generated using the the notebook [create_datasets](../../data/create_datasets.ipynb) in the [data](../../data) directory in the root directory of this repo for instruction to download the following dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We load the Quandl stock price datasets covering the US equity markets 2000-18 using `pd.IndexSlice` to perform a slice operation on the `pd.MultiIndex`, select the adjusted close price and unpivot the column to convert the DataFrame to wide format with tickers in the columns and timestamps in the rows:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set data store location:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:29:18.142415Z",
     "start_time": "2021-02-23T20:29:18.135503Z"
    }
   },
   "outputs": [],
   "source": [
    "DATA_STORE = '../data/assets.h5'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:29:18.154819Z",
     "start_time": "2021-02-23T20:29:18.143752Z"
    }
   },
   "outputs": [],
   "source": [
    "YEAR = 12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:29:18.164476Z",
     "start_time": "2021-02-23T20:29:18.155706Z"
    }
   },
   "outputs": [],
   "source": [
    "START = 1995\n",
    "END = 2017"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:29:24.397333Z",
     "start_time": "2021-02-23T20:29:18.165527Z"
    }
   },
   "outputs": [],
   "source": [
    "with pd.HDFStore(DATA_STORE) as store:\n",
    "    prices = (store['quandl/wiki/prices']\n",
    "              .loc[idx[str(START):str(END), :], :]\n",
    "              .filter(like='adj_')\n",
    "              .dropna()\n",
    "              .swaplevel()\n",
    "              .rename(columns=lambda x: x.replace('adj_', ''))\n",
    "              .join(store['us_equities/stocks']\n",
    "                    .loc[:, ['sector']])\n",
    "              .dropna())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:29:24.888343Z",
     "start_time": "2021-02-23T20:29:24.398493Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 10241831 entries, ('AAN', Timestamp('1995-01-03 00:00:00')) to ('ZUMZ', Timestamp('2017-12-29 00:00:00'))\n",
      "Data columns (total 6 columns):\n",
      " #   Column  Non-Null Count     Dtype  \n",
      "---  ------  --------------     -----  \n",
      " 0   open    10241831 non-null  float64\n",
      " 1   high    10241831 non-null  float64\n",
      " 2   low     10241831 non-null  float64\n",
      " 3   close   10241831 non-null  float64\n",
      " 4   volume  10241831 non-null  float64\n",
      " 5   sector  10241831 non-null  object \n",
      "dtypes: float64(5), object(1)\n",
      "memory usage: 508.7+ MB\n"
     ]
    }
   ],
   "source": [
    "prices.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:29:24.974142Z",
     "start_time": "2021-02-23T20:29:24.890337Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2369"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(prices.index.unique('ticker'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Remove stocks with less than ten years of data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:30:10.981706Z",
     "start_time": "2021-02-23T20:29:24.975850Z"
    }
   },
   "outputs": [],
   "source": [
    "min_obs = 10 * 252\n",
    "nobs = prices.groupby(level='ticker').size()\n",
    "to_drop = nobs[nobs < min_obs].index\n",
    "prices = prices.drop(to_drop, level='ticker')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:30:11.434460Z",
     "start_time": "2021-02-23T20:30:10.982824Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 9532628 entries, ('AAN', Timestamp('1995-01-03 00:00:00')) to ('ZUMZ', Timestamp('2017-12-29 00:00:00'))\n",
      "Data columns (total 6 columns):\n",
      " #   Column  Non-Null Count    Dtype  \n",
      "---  ------  --------------    -----  \n",
      " 0   open    9532628 non-null  float64\n",
      " 1   high    9532628 non-null  float64\n",
      " 2   low     9532628 non-null  float64\n",
      " 3   close   9532628 non-null  float64\n",
      " 4   volume  9532628 non-null  float64\n",
      " 5   sector  9532628 non-null  object \n",
      "dtypes: float64(5), object(1)\n",
      "memory usage: 473.5+ MB\n"
     ]
    }
   ],
   "source": [
    "prices.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:30:11.509683Z",
     "start_time": "2021-02-23T20:30:11.435475Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1883"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(prices.index.unique('ticker'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Add some Basic Factors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute the Relative Strength Index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:30:58.402376Z",
     "start_time": "2021-02-23T20:30:11.510780Z"
    }
   },
   "outputs": [],
   "source": [
    "prices['rsi'] = prices.groupby(level='ticker').close.apply(RSI)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:31:21.858293Z",
     "start_time": "2021-02-23T20:30:58.403322Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(prices.rsi);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute Bollinger Bands"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:31:21.862142Z",
     "start_time": "2021-02-23T20:31:21.859408Z"
    }
   },
   "outputs": [],
   "source": [
    "def compute_bb(close):\n",
    "    high, mid, low = BBANDS(np.log1p(close), timeperiod=20)\n",
    "    return pd.DataFrame({'bb_high': high,\n",
    "                         'bb_mid': mid,\n",
    "                         'bb_low': low}, index=close.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:32:27.635722Z",
     "start_time": "2021-02-23T20:31:21.863546Z"
    }
   },
   "outputs": [],
   "source": [
    "prices = (prices.join(prices\n",
    "                      .groupby(level='ticker')\n",
    "                      .close\n",
    "                      .apply(compute_bb)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:32:28.025953Z",
     "start_time": "2021-02-23T20:32:27.636648Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 9532628 entries, ('AAN', Timestamp('1995-01-03 00:00:00')) to ('ZUMZ', Timestamp('2017-12-29 00:00:00'))\n",
      "Data columns (total 10 columns):\n",
      " #   Column   Non-Null Count    Dtype  \n",
      "---  ------   --------------    -----  \n",
      " 0   open     9532628 non-null  float64\n",
      " 1   high     9532628 non-null  float64\n",
      " 2   low      9532628 non-null  float64\n",
      " 3   close    9532628 non-null  float64\n",
      " 4   volume   9532628 non-null  float64\n",
      " 5   sector   9532628 non-null  object \n",
      " 6   rsi      9506266 non-null  float64\n",
      " 7   bb_high  9496851 non-null  float64\n",
      " 8   bb_mid   9496851 non-null  float64\n",
      " 9   bb_low   9496851 non-null  float64\n",
      "dtypes: float64(9), object(1)\n",
      "memory usage: 1022.4+ MB\n"
     ]
    }
   ],
   "source": [
    "prices.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:32:29.422206Z",
     "start_time": "2021-02-23T20:32:28.026886Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>bb_high</th>\n",
       "      <th>bb_mid</th>\n",
       "      <th>bb_low</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>9.496851e+06</td>\n",
       "      <td>9.496851e+06</td>\n",
       "      <td>9.496851e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.954140e+00</td>\n",
       "      <td>2.881157e+00</td>\n",
       "      <td>2.808174e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.024536e+00</td>\n",
       "      <td>1.026901e+00</td>\n",
       "      <td>1.032999e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>8.933146e-03</td>\n",
       "      <td>8.933146e-03</td>\n",
       "      <td>-1.568426e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.303724e+00</td>\n",
       "      <td>2.226078e+00</td>\n",
       "      <td>2.146471e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.940911e+00</td>\n",
       "      <td>2.868116e+00</td>\n",
       "      <td>2.796484e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.555602e+00</td>\n",
       "      <td>3.487039e+00</td>\n",
       "      <td>3.420498e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.376991e+01</td>\n",
       "      <td>1.358056e+01</td>\n",
       "      <td>1.346225e+01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            bb_high        bb_mid        bb_low\n",
       "count  9.496851e+06  9.496851e+06  9.496851e+06\n",
       "mean   2.954140e+00  2.881157e+00  2.808174e+00\n",
       "std    1.024536e+00  1.026901e+00  1.032999e+00\n",
       "min    8.933146e-03  8.933146e-03 -1.568426e+00\n",
       "25%    2.303724e+00  2.226078e+00  2.146471e+00\n",
       "50%    2.940911e+00  2.868116e+00  2.796484e+00\n",
       "75%    3.555602e+00  3.487039e+00  3.420498e+00\n",
       "max    1.376991e+01  1.358056e+01  1.346225e+01"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prices.filter(like='bb_').describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:33:37.302678Z",
     "start_time": "2021-02-23T20:32:29.423220Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(ncols=3, figsize=(15,4))\n",
    "for i, col in enumerate(['bb_low', 'bb_mid', 'bb_low']):\n",
    "    sns.distplot(prices[col], ax=axes[i])\n",
    "    axes[i].set_title(col);\n",
    "fig.tight_layout();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:33:37.614610Z",
     "start_time": "2021-02-23T20:33:37.303801Z"
    }
   },
   "outputs": [],
   "source": [
    "prices['bb_up'] = prices.bb_high.sub(np.log1p(prices.close))\n",
    "prices['bb_down'] = np.log1p(prices.close).sub(prices.bb_low)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:33:42.626284Z",
     "start_time": "2021-02-23T20:33:37.615839Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(ncols=2, figsize=(10,4))\n",
    "for i, col in enumerate(['bb_down', 'bb_up']):\n",
    "    sns.boxenplot(prices[col], ax=axes[i])\n",
    "    axes[i].set_title(col);\n",
    "fig.tight_layout();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute Average True Range"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Helper for indicators with multiple inputs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:33:42.827296Z",
     "start_time": "2021-02-23T20:33:42.627221Z"
    }
   },
   "outputs": [],
   "source": [
    "by_ticker = prices.groupby('ticker', group_keys=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:33:42.831678Z",
     "start_time": "2021-02-23T20:33:42.828250Z"
    }
   },
   "outputs": [],
   "source": [
    "def compute_atr(stock_data):\n",
    "    atr = ATR(stock_data.high, \n",
    "              stock_data.low, \n",
    "              stock_data.close, \n",
    "              timeperiod=14)\n",
    "    return atr.sub(atr.mean()).div(atr.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:34:04.342244Z",
     "start_time": "2021-02-23T20:33:42.833588Z"
    }
   },
   "outputs": [],
   "source": [
    "prices['atr'] = by_ticker.apply(compute_atr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:34:27.080212Z",
     "start_time": "2021-02-23T20:34:04.343606Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(prices.atr);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:34:48.557533Z",
     "start_time": "2021-02-23T20:34:27.081545Z"
    }
   },
   "outputs": [],
   "source": [
    "prices['natr'] = by_ticker.apply(lambda x: NATR(high=x.high, low=x.low, close=x.close))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:35:12.618207Z",
     "start_time": "2021-02-23T20:34:48.558475Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(prices.natr[prices.natr<10]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute Moving Average Convergence/Divergence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:07.366887Z",
     "start_time": "2021-02-23T20:35:12.619395Z"
    },
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "def compute_macd(close):\n",
    "    macd = MACD(close)[0]\n",
    "    return macd.sub(macd.mean()).div(macd.std())\n",
    "\n",
    "prices['macd'] = prices.groupby(level='ticker').close.apply(compute_macd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:31.231218Z",
     "start_time": "2021-02-23T20:36:07.368084Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(prices.macd);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compute dollar volume to determine universe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:31.308363Z",
     "start_time": "2021-02-23T20:36:31.232479Z"
    }
   },
   "outputs": [],
   "source": [
    "prices['dollar_volume'] = (prices.loc[:, 'close']\n",
    "                           .mul(prices.loc[:, 'volume'], axis=0))\n",
    "\n",
    "prices.dollar_volume /= 1e6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:32.915677Z",
     "start_time": "2021-02-23T20:36:31.309455Z"
    }
   },
   "outputs": [],
   "source": [
    "prices.to_hdf('data.h5', 'us/equities/prices')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:36.371266Z",
     "start_time": "2021-02-23T20:36:32.917114Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 9532628 entries, ('AAN', Timestamp('1995-01-03 00:00:00')) to ('ZUMZ', Timestamp('2017-12-29 00:00:00'))\n",
      "Data columns (total 16 columns):\n",
      " #   Column         Non-Null Count    Dtype  \n",
      "---  ------         --------------    -----  \n",
      " 0   open           9532628 non-null  float64\n",
      " 1   high           9532628 non-null  float64\n",
      " 2   low            9532628 non-null  float64\n",
      " 3   close          9532628 non-null  float64\n",
      " 4   volume         9532628 non-null  float64\n",
      " 5   sector         9532628 non-null  object \n",
      " 6   rsi            9506266 non-null  float64\n",
      " 7   bb_high        9496851 non-null  float64\n",
      " 8   bb_mid         9496851 non-null  float64\n",
      " 9   bb_low         9496851 non-null  float64\n",
      " 10  bb_up          9496851 non-null  float64\n",
      " 11  bb_down        9496851 non-null  float64\n",
      " 12  atr            9506266 non-null  float64\n",
      " 13  natr           9506266 non-null  float64\n",
      " 14  macd           9470489 non-null  float64\n",
      " 15  dollar_volume  9532628 non-null  float64\n",
      "dtypes: float64(15), object(1)\n",
      "memory usage: 1.2+ GB\n"
     ]
    }
   ],
   "source": [
    "prices = pd.read_hdf('data.h5', 'us/equities/prices')\n",
    "prices.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resample OHLCV prices to monthly frequency"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To reduce training time and experiment with strategies for longer time horizons, we convert the business-daily data to month-end frequency using the available adjusted close price:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:36.374719Z",
     "start_time": "2021-02-23T20:36:36.372390Z"
    }
   },
   "outputs": [],
   "source": [
    "last_cols = [c for c in prices.columns.unique(0) if c not in ['dollar_volume', 'volume',\n",
    "                                                              'open', 'high', 'low']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:39.763558Z",
     "start_time": "2021-02-23T20:36:36.375717Z"
    }
   },
   "outputs": [],
   "source": [
    "prices = prices.unstack('ticker')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:46.968273Z",
     "start_time": "2021-02-23T20:36:39.764800Z"
    }
   },
   "outputs": [],
   "source": [
    "data = (pd.concat([prices.dollar_volume.resample('M').mean().stack('ticker').to_frame('dollar_volume'),\n",
    "                   prices[last_cols].resample('M').last().stack('ticker')],\n",
    "                  axis=1)\n",
    "        .swaplevel()\n",
    "        .dropna())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:36:47.009662Z",
     "start_time": "2021-02-23T20:36:46.969428Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 452529 entries, ('AAN', Timestamp('1995-02-28 00:00:00')) to ('ZUMZ', Timestamp('2017-12-31 00:00:00'))\n",
      "Data columns (total 12 columns):\n",
      " #   Column         Non-Null Count   Dtype  \n",
      "---  ------         --------------   -----  \n",
      " 0   dollar_volume  452529 non-null  float64\n",
      " 1   atr            452529 non-null  float64\n",
      " 2   bb_down        452529 non-null  float64\n",
      " 3   bb_high        452529 non-null  float64\n",
      " 4   bb_low         452529 non-null  float64\n",
      " 5   bb_mid         452529 non-null  float64\n",
      " 6   bb_up          452529 non-null  float64\n",
      " 7   close          452529 non-null  float64\n",
      " 8   macd           452529 non-null  float64\n",
      " 9   natr           452529 non-null  float64\n",
      " 10  rsi            452529 non-null  float64\n",
      " 11  sector         452529 non-null  object \n",
      "dtypes: float64(11), object(1)\n",
      "memory usage: 43.2+ MB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Select 500 most-traded equities"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Select the 500 most-traded stocks based on a 5-year rolling average of dollar volume."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:01.705623Z",
     "start_time": "2021-02-23T20:40:58.507121Z"
    }
   },
   "outputs": [],
   "source": [
    "data['dollar_volume'] = (data.loc[:, 'dollar_volume']\n",
    "                         .unstack('ticker')\n",
    "                         .rolling(window=5*12, min_periods=12)\n",
    "                         .mean()\n",
    "                         .stack()\n",
    "                         .swaplevel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:01.887279Z",
     "start_time": "2021-02-23T20:41:01.740521Z"
    }
   },
   "outputs": [],
   "source": [
    "data['dollar_vol_rank'] = (data\n",
    "                           .groupby('date')\n",
    "                           .dollar_volume\n",
    "                           .rank(ascending=False))\n",
    "\n",
    "data = data[data.dollar_vol_rank < 500].drop(['dollar_volume', 'dollar_vol_rank'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:01.892527Z",
     "start_time": "2021-02-23T20:41:01.888455Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "905"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data.index.unique('ticker'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create monthly return series"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To capture time series dynamics that reflect, for example, momentum patterns, we compute historical returns using the method `.pct_change(n_periods)`, that is, returns over various monthly periods as identified by lags.\n",
    "\n",
    "We then convert the wide result back to long format with the `.stack()` method, use `.pipe()` to apply the `.clip()` method to the resulting `DataFrame`, and winsorize returns at the [1%, 99%] levels; that is, we cap outliers at these percentiles.\n",
    "\n",
    "Finally, we normalize returns using the geometric average. After using `.swaplevel()` to change the order of the `MultiIndex` levels, we obtain compounded monthly returns for six periods ranging from 1 to 12 months:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:01.905281Z",
     "start_time": "2021-02-23T20:41:01.894367Z"
    }
   },
   "outputs": [],
   "source": [
    "outlier_cutoff = 0.01\n",
    "lags = [1, 3, 6, 12]\n",
    "returns = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:07.775321Z",
     "start_time": "2021-02-23T20:41:01.906659Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 202879 entries, ('AAPL', Timestamp('1996-02-29 00:00:00')) to ('ZIXI', Timestamp('2017-12-31 00:00:00'))\n",
      "Data columns (total 4 columns):\n",
      " #   Column      Non-Null Count   Dtype  \n",
      "---  ------      --------------   -----  \n",
      " 0   return_1m   202879 non-null  float64\n",
      " 1   return_3m   201069 non-null  float64\n",
      " 2   return_6m   198362 non-null  float64\n",
      " 3   return_12m  192972 non-null  float64\n",
      "dtypes: float64(4)\n",
      "memory usage: 7.0+ MB\n"
     ]
    }
   ],
   "source": [
    "for lag in lags:\n",
    "    returns.append(data\n",
    "                   .close\n",
    "                   .unstack('ticker')\n",
    "                   .sort_index()\n",
    "                   .pct_change(lag)\n",
    "                   .stack('ticker')\n",
    "                   .pipe(lambda x: x.clip(lower=x.quantile(outlier_cutoff),\n",
    "                                          upper=x.quantile(1-outlier_cutoff)))\n",
    "                   .add(1)\n",
    "                   .pow(1/lag)\n",
    "                   .sub(1)\n",
    "                   .to_frame(f'return_{lag}m')\n",
    "                   )\n",
    "    \n",
    "returns = pd.concat(returns, axis=1).swaplevel()\n",
    "returns.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:11.269318Z",
     "start_time": "2021-02-23T20:41:11.221582Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>return_1m</th>\n",
       "      <th>return_3m</th>\n",
       "      <th>return_6m</th>\n",
       "      <th>return_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>202879.000000</td>\n",
       "      <td>201069.000000</td>\n",
       "      <td>198362.000000</td>\n",
       "      <td>192972.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.007333</td>\n",
       "      <td>0.004950</td>\n",
       "      <td>0.004345</td>\n",
       "      <td>0.004167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.085992</td>\n",
       "      <td>0.050986</td>\n",
       "      <td>0.037222</td>\n",
       "      <td>0.027246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.273767</td>\n",
       "      <td>-0.179499</td>\n",
       "      <td>-0.134177</td>\n",
       "      <td>-0.093876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.011852</td>\n",
       "      <td>-0.003188</td>\n",
       "      <td>-0.000486</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.032949</td>\n",
       "      <td>0.023727</td>\n",
       "      <td>0.018920</td>\n",
       "      <td>0.015461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.331175</td>\n",
       "      <td>0.180733</td>\n",
       "      <td>0.131169</td>\n",
       "      <td>0.099726</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           return_1m      return_3m      return_6m     return_12m\n",
       "count  202879.000000  201069.000000  198362.000000  192972.000000\n",
       "mean        0.007333       0.004950       0.004345       0.004167\n",
       "std         0.085992       0.050986       0.037222       0.027246\n",
       "min        -0.273767      -0.179499      -0.134177      -0.093876\n",
       "25%        -0.011852      -0.003188      -0.000486       0.000000\n",
       "50%         0.000000       0.000000       0.000000       0.000000\n",
       "75%         0.032949       0.023727       0.018920       0.015461\n",
       "max         0.331175       0.180733       0.131169       0.099726"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "returns.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:14.718336Z",
     "start_time": "2021-02-23T20:41:14.046696Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmap = sns.diverging_palette(10, 220, as_cmap=True)\n",
    "sns.clustermap(returns.corr('spearman'), annot=True, center=0, cmap=cmap);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:18.324087Z",
     "start_time": "2021-02-23T20:41:17.783552Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 121589 entries, ('AAPL', Timestamp('1997-01-31 00:00:00')) to ('ZION', Timestamp('2017-12-31 00:00:00'))\n",
      "Data columns (total 14 columns):\n",
      " #   Column      Non-Null Count   Dtype  \n",
      "---  ------      --------------   -----  \n",
      " 0   atr         121589 non-null  float64\n",
      " 1   bb_down     121589 non-null  float64\n",
      " 2   bb_high     121589 non-null  float64\n",
      " 3   bb_low      121589 non-null  float64\n",
      " 4   bb_mid      121589 non-null  float64\n",
      " 5   bb_up       121589 non-null  float64\n",
      " 6   macd        121589 non-null  float64\n",
      " 7   natr        121589 non-null  float64\n",
      " 8   rsi         121589 non-null  float64\n",
      " 9   sector      121589 non-null  object \n",
      " 10  return_1m   121589 non-null  float64\n",
      " 11  return_3m   121589 non-null  float64\n",
      " 12  return_6m   121589 non-null  float64\n",
      " 13  return_12m  121589 non-null  float64\n",
      "dtypes: float64(13), object(1)\n",
      "memory usage: 13.5+ MB\n"
     ]
    }
   ],
   "source": [
    "data = data.join(returns).drop('close', axis=1).dropna()\n",
    "data.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:20.013017Z",
     "start_time": "2021-02-23T20:41:18.325557Z"
    }
   },
   "outputs": [],
   "source": [
    "min_obs = 5*12\n",
    "nobs = data.groupby(level='ticker').size()\n",
    "to_drop = nobs[nobs < min_obs].index\n",
    "data = data.drop(to_drop, level='ticker')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:41:20.018610Z",
     "start_time": "2021-02-23T20:41:20.014356Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "613"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data.index.unique('ticker'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are left with 613 tickers."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rolling Factor Betas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will introduce the Fama—French data to estimate the exposure of assets to common risk factors using linear regression in [Chapter 8, Time Series Models]([](../../08_time_series_models))."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The five Fama—French factors, namely market risk, size, value, operating profitability, and investment have been shown empirically to explain asset returns and are commonly used to assess the risk/return profile of portfolios. Hence, it is natural to include past factor exposures as financial features in models that aim to predict future returns."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can access the historical factor returns using the `pandas-datareader` and estimate historical exposures using the `PandasRollingOLS` rolling linear regression functionality in the `pyfinance` library as follows:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use Fama-French research factors to estimate the factor exposures of the stock in the dataset to the 5 factors market risk, size, value, operating profitability and investment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:11.097763Z",
     "start_time": "2021-02-23T20:42:10.880529Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 314 entries, 1995-01-31 to 2021-02-28\n",
      "Freq: M\n",
      "Data columns (total 5 columns):\n",
      " #   Column  Non-Null Count  Dtype  \n",
      "---  ------  --------------  -----  \n",
      " 0   Mkt-RF  314 non-null    float64\n",
      " 1   SMB     314 non-null    float64\n",
      " 2   HML     314 non-null    float64\n",
      " 3   RMW     314 non-null    float64\n",
      " 4   CMA     314 non-null    float64\n",
      "dtypes: float64(5)\n",
      "memory usage: 14.7 KB\n"
     ]
    }
   ],
   "source": [
    "factors = ['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA']\n",
    "factor_data = web.DataReader('F-F_Research_Data_5_Factors_2x3', \n",
    "                             'famafrench', \n",
    "                             start=START)[0].drop('RF', axis=1)\n",
    "factor_data.index = factor_data.index.to_timestamp()\n",
    "factor_data = factor_data.resample('M').last().div(100)\n",
    "factor_data.index.name = 'date'\n",
    "factor_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:12.481069Z",
     "start_time": "2021-02-23T20:42:12.417368Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 115181 entries, ('A', Timestamp('2001-12-31 00:00:00', freq='M')) to ('ZION', Timestamp('2017-12-31 00:00:00', freq='M'))\n",
      "Data columns (total 6 columns):\n",
      " #   Column     Non-Null Count   Dtype  \n",
      "---  ------     --------------   -----  \n",
      " 0   Mkt-RF     115181 non-null  float64\n",
      " 1   SMB        115181 non-null  float64\n",
      " 2   HML        115181 non-null  float64\n",
      " 3   RMW        115181 non-null  float64\n",
      " 4   CMA        115181 non-null  float64\n",
      " 5   return_1m  115181 non-null  float64\n",
      "dtypes: float64(6)\n",
      "memory usage: 5.8+ MB\n"
     ]
    }
   ],
   "source": [
    "factor_data = factor_data.join(data['return_1m']).dropna().sort_index()\n",
    "factor_data['return_1m'] -= factor_data['Mkt-RF']\n",
    "factor_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:12.761795Z",
     "start_time": "2021-02-23T20:42:12.722981Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Mkt-RF</th>\n",
       "      <th>SMB</th>\n",
       "      <th>HML</th>\n",
       "      <th>RMW</th>\n",
       "      <th>CMA</th>\n",
       "      <th>return_1m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>115181.000000</td>\n",
       "      <td>115181.000000</td>\n",
       "      <td>115181.000000</td>\n",
       "      <td>115181.000000</td>\n",
       "      <td>115181.00000</td>\n",
       "      <td>115181.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.006184</td>\n",
       "      <td>0.002261</td>\n",
       "      <td>0.001918</td>\n",
       "      <td>0.003343</td>\n",
       "      <td>0.00242</td>\n",
       "      <td>0.005456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.044295</td>\n",
       "      <td>0.031484</td>\n",
       "      <td>0.031217</td>\n",
       "      <td>0.029219</td>\n",
       "      <td>0.02146</td>\n",
       "      <td>0.092551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.172300</td>\n",
       "      <td>-0.148900</td>\n",
       "      <td>-0.111200</td>\n",
       "      <td>-0.184800</td>\n",
       "      <td>-0.06860</td>\n",
       "      <td>-0.387267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.019700</td>\n",
       "      <td>-0.016900</td>\n",
       "      <td>-0.015100</td>\n",
       "      <td>-0.011600</td>\n",
       "      <td>-0.01060</td>\n",
       "      <td>-0.043902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.011700</td>\n",
       "      <td>0.001100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.004200</td>\n",
       "      <td>-0.00020</td>\n",
       "      <td>0.002598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.034900</td>\n",
       "      <td>0.022700</td>\n",
       "      <td>0.018000</td>\n",
       "      <td>0.014700</td>\n",
       "      <td>0.01430</td>\n",
       "      <td>0.050718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.113500</td>\n",
       "      <td>0.180800</td>\n",
       "      <td>0.125800</td>\n",
       "      <td>0.133800</td>\n",
       "      <td>0.09560</td>\n",
       "      <td>0.503475</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Mkt-RF            SMB            HML            RMW  \\\n",
       "count  115181.000000  115181.000000  115181.000000  115181.000000   \n",
       "mean        0.006184       0.002261       0.001918       0.003343   \n",
       "std         0.044295       0.031484       0.031217       0.029219   \n",
       "min        -0.172300      -0.148900      -0.111200      -0.184800   \n",
       "25%        -0.019700      -0.016900      -0.015100      -0.011600   \n",
       "50%         0.011700       0.001100       0.000200       0.004200   \n",
       "75%         0.034900       0.022700       0.018000       0.014700   \n",
       "max         0.113500       0.180800       0.125800       0.133800   \n",
       "\n",
       "                CMA      return_1m  \n",
       "count  115181.00000  115181.000000  \n",
       "mean        0.00242       0.005456  \n",
       "std         0.02146       0.092551  \n",
       "min        -0.06860      -0.387267  \n",
       "25%        -0.01060      -0.043902  \n",
       "50%        -0.00020       0.002598  \n",
       "75%         0.01430       0.050718  \n",
       "max         0.09560       0.503475  "
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "factor_data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:17.214895Z",
     "start_time": "2021-02-23T20:42:15.331075Z"
    }
   },
   "outputs": [],
   "source": [
    "T = 60\n",
    "# betas = (factor_data\n",
    "#          .groupby(level='ticker', group_keys=False)\n",
    "#          .apply(lambda x: PandasRollingOLS(window=min(T, x.shape[0]-1), \n",
    "#                                            y=x.return_1m, \n",
    "#                                            x=x.drop('return_1m', axis=1)).beta)\n",
    "#         .rename(columns={'Mkt-RF': 'beta'}))\n",
    "betas = (factor_data.groupby(level='ticker',\n",
    "                             group_keys=False)\n",
    "         .apply(lambda x: RollingOLS(endog=x.return_1m,\n",
    "                                     exog=sm.add_constant(x.drop('return_1m', axis=1)),\n",
    "                                     window=min(T, x.shape[0]-1))\n",
    "                .fit(params_only=True)\n",
    "                .params\n",
    "                .rename(columns={'Mkt-RF': 'beta'})\n",
    "                .drop('const', axis=1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>beta</th>\n",
       "      <th>SMB</th>\n",
       "      <th>HML</th>\n",
       "      <th>RMW</th>\n",
       "      <th>CMA</th>\n",
       "      <th>total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>115181.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.067841</td>\n",
       "      <td>0.192202</td>\n",
       "      <td>0.116937</td>\n",
       "      <td>-0.012575</td>\n",
       "      <td>0.004220</td>\n",
       "      <td>0.252877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.477380</td>\n",
       "      <td>0.587193</td>\n",
       "      <td>0.821477</td>\n",
       "      <td>0.877605</td>\n",
       "      <td>0.965009</td>\n",
       "      <td>1.377355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-1.825339</td>\n",
       "      <td>-1.960546</td>\n",
       "      <td>-4.056384</td>\n",
       "      <td>-5.224988</td>\n",
       "      <td>-5.204906</td>\n",
       "      <td>-10.257558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.252872</td>\n",
       "      <td>-0.204076</td>\n",
       "      <td>-0.411242</td>\n",
       "      <td>-0.492922</td>\n",
       "      <td>-0.536780</td>\n",
       "      <td>-0.135471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.047046</td>\n",
       "      <td>0.151163</td>\n",
       "      <td>0.059990</td>\n",
       "      <td>0.052307</td>\n",
       "      <td>0.053130</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.370298</td>\n",
       "      <td>0.546711</td>\n",
       "      <td>0.593241</td>\n",
       "      <td>0.543717</td>\n",
       "      <td>0.596435</td>\n",
       "      <td>0.842115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.688829</td>\n",
       "      <td>3.286469</td>\n",
       "      <td>4.716294</td>\n",
       "      <td>4.029144</td>\n",
       "      <td>5.129094</td>\n",
       "      <td>10.580404</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               beta           SMB           HML           RMW           CMA  \\\n",
       "count  79014.000000  79014.000000  79014.000000  79014.000000  79014.000000   \n",
       "mean       0.067841      0.192202      0.116937     -0.012575      0.004220   \n",
       "std        0.477380      0.587193      0.821477      0.877605      0.965009   \n",
       "min       -1.825339     -1.960546     -4.056384     -5.224988     -5.204906   \n",
       "25%       -0.252872     -0.204076     -0.411242     -0.492922     -0.536780   \n",
       "50%        0.047046      0.151163      0.059990      0.052307      0.053130   \n",
       "75%        0.370298      0.546711      0.593241      0.543717      0.596435   \n",
       "max        2.688829      3.286469      4.716294      4.029144      5.129094   \n",
       "\n",
       "               total  \n",
       "count  115181.000000  \n",
       "mean        0.252877  \n",
       "std         1.377355  \n",
       "min       -10.257558  \n",
       "25%        -0.135471  \n",
       "50%         0.000000  \n",
       "75%         0.842115  \n",
       "max        10.580404  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "betas.describe().join(betas.sum(1).describe().to_frame('total'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:17.251480Z",
     "start_time": "2021-02-23T20:42:17.216084Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>beta</th>\n",
       "      <th>SMB</th>\n",
       "      <th>HML</th>\n",
       "      <th>RMW</th>\n",
       "      <th>CMA</th>\n",
       "      <th>total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>79014.000000</td>\n",
       "      <td>115181.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.067841</td>\n",
       "      <td>0.192202</td>\n",
       "      <td>0.116937</td>\n",
       "      <td>-0.012575</td>\n",
       "      <td>0.004220</td>\n",
       "      <td>0.252877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.477380</td>\n",
       "      <td>0.587193</td>\n",
       "      <td>0.821477</td>\n",
       "      <td>0.877605</td>\n",
       "      <td>0.965009</td>\n",
       "      <td>1.377355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-1.825339</td>\n",
       "      <td>-1.960546</td>\n",
       "      <td>-4.056384</td>\n",
       "      <td>-5.224988</td>\n",
       "      <td>-5.204906</td>\n",
       "      <td>-10.257558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.252872</td>\n",
       "      <td>-0.204076</td>\n",
       "      <td>-0.411242</td>\n",
       "      <td>-0.492922</td>\n",
       "      <td>-0.536780</td>\n",
       "      <td>-0.135471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.047046</td>\n",
       "      <td>0.151163</td>\n",
       "      <td>0.059990</td>\n",
       "      <td>0.052307</td>\n",
       "      <td>0.053130</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.370298</td>\n",
       "      <td>0.546711</td>\n",
       "      <td>0.593241</td>\n",
       "      <td>0.543717</td>\n",
       "      <td>0.596435</td>\n",
       "      <td>0.842115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.688829</td>\n",
       "      <td>3.286469</td>\n",
       "      <td>4.716294</td>\n",
       "      <td>4.029144</td>\n",
       "      <td>5.129094</td>\n",
       "      <td>10.580404</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               beta           SMB           HML           RMW           CMA  \\\n",
       "count  79014.000000  79014.000000  79014.000000  79014.000000  79014.000000   \n",
       "mean       0.067841      0.192202      0.116937     -0.012575      0.004220   \n",
       "std        0.477380      0.587193      0.821477      0.877605      0.965009   \n",
       "min       -1.825339     -1.960546     -4.056384     -5.224988     -5.204906   \n",
       "25%       -0.252872     -0.204076     -0.411242     -0.492922     -0.536780   \n",
       "50%        0.047046      0.151163      0.059990      0.052307      0.053130   \n",
       "75%        0.370298      0.546711      0.593241      0.543717      0.596435   \n",
       "max        2.688829      3.286469      4.716294      4.029144      5.129094   \n",
       "\n",
       "               total  \n",
       "count  115181.000000  \n",
       "mean        0.252877  \n",
       "std         1.377355  \n",
       "min       -10.257558  \n",
       "25%        -0.135471  \n",
       "50%         0.000000  \n",
       "75%         0.842115  \n",
       "max        10.580404  "
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "betas.describe().join(betas.sum(1).describe().to_frame('total'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:17.624730Z",
     "start_time": "2021-02-23T20:42:17.252923Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmap = sns.diverging_palette(10, 220, as_cmap=True)\n",
    "sns.clustermap(betas.corr(), annot=True, cmap=cmap, center=0);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:20.852056Z",
     "start_time": "2021-02-23T20:42:20.613962Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "data = (data\n",
    "        .join(betas\n",
    "              .groupby(level='ticker')\n",
    "              .shift())\n",
    "       .dropna()\n",
    "       .sort_index())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:21.019345Z",
     "start_time": "2021-02-23T20:42:21.005378Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 78401 entries, ('A', Timestamp('2006-12-31 00:00:00')) to ('ZION', Timestamp('2017-12-31 00:00:00'))\n",
      "Data columns (total 19 columns):\n",
      " #   Column      Non-Null Count  Dtype  \n",
      "---  ------      --------------  -----  \n",
      " 0   atr         78401 non-null  float64\n",
      " 1   bb_down     78401 non-null  float64\n",
      " 2   bb_high     78401 non-null  float64\n",
      " 3   bb_low      78401 non-null  float64\n",
      " 4   bb_mid      78401 non-null  float64\n",
      " 5   bb_up       78401 non-null  float64\n",
      " 6   macd        78401 non-null  float64\n",
      " 7   natr        78401 non-null  float64\n",
      " 8   rsi         78401 non-null  float64\n",
      " 9   sector      78401 non-null  object \n",
      " 10  return_1m   78401 non-null  float64\n",
      " 11  return_3m   78401 non-null  float64\n",
      " 12  return_6m   78401 non-null  float64\n",
      " 13  return_12m  78401 non-null  float64\n",
      " 14  beta        78401 non-null  float64\n",
      " 15  SMB         78401 non-null  float64\n",
      " 16  HML         78401 non-null  float64\n",
      " 17  RMW         78401 non-null  float64\n",
      " 18  CMA         78401 non-null  float64\n",
      "dtypes: float64(18), object(1)\n",
      "memory usage: 11.8+ MB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Momentum factors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use these results to compute momentum factors based on the difference between returns over longer periods and the most recent monthly return, as well as for the difference between 3 and 12 month returns as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:23.799079Z",
     "start_time": "2021-02-23T20:42:23.793520Z"
    }
   },
   "outputs": [],
   "source": [
    "for lag in [3, 6, 12]:\n",
    "    data[f'momentum_{lag}'] = data[f'return_{lag}m'].sub(data.return_1m)\n",
    "    if lag > 3:\n",
    "        data[f'momentum_3_{lag}'] = data[f'return_{lag}m'].sub(data.return_3m)    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Date Indicators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:24.107325Z",
     "start_time": "2021-02-23T20:42:24.092809Z"
    }
   },
   "outputs": [],
   "source": [
    "dates = data.index.get_level_values('date')\n",
    "data['year'] = dates.year\n",
    "data['month'] = dates.month"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Target: Holding Period Returns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To compute returns for our one-month target holding period, we use the returns computed previously and shift them back to align them with the current financial features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:26.197129Z",
     "start_time": "2021-02-23T20:42:26.187948Z"
    }
   },
   "outputs": [],
   "source": [
    "data['target'] = data.groupby(level='ticker')[f'return_1m'].shift(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:26.392027Z",
     "start_time": "2021-02-23T20:42:26.374282Z"
    }
   },
   "outputs": [],
   "source": [
    "data = data.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:26.573080Z",
     "start_time": "2021-02-23T20:42:26.548680Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 77788 entries, ('A', Timestamp('2006-12-31 00:00:00')) to ('ZION', Timestamp('2017-11-30 00:00:00'))\n",
      "Data columns (total 27 columns):\n",
      " #   Column         Non-Null Count  Dtype  \n",
      "---  ------         --------------  -----  \n",
      " 0   atr            77788 non-null  float64\n",
      " 1   bb_down        77788 non-null  float64\n",
      " 2   bb_high        77788 non-null  float64\n",
      " 3   bb_low         77788 non-null  float64\n",
      " 4   bb_mid         77788 non-null  float64\n",
      " 5   bb_up          77788 non-null  float64\n",
      " 6   macd           77788 non-null  float64\n",
      " 7   natr           77788 non-null  float64\n",
      " 8   rsi            77788 non-null  float64\n",
      " 9   sector         77788 non-null  object \n",
      " 10  return_1m      77788 non-null  float64\n",
      " 11  return_3m      77788 non-null  float64\n",
      " 12  return_6m      77788 non-null  float64\n",
      " 13  return_12m     77788 non-null  float64\n",
      " 14  beta           77788 non-null  float64\n",
      " 15  SMB            77788 non-null  float64\n",
      " 16  HML            77788 non-null  float64\n",
      " 17  RMW            77788 non-null  float64\n",
      " 18  CMA            77788 non-null  float64\n",
      " 19  momentum_3     77788 non-null  float64\n",
      " 20  momentum_6     77788 non-null  float64\n",
      " 21  momentum_3_6   77788 non-null  float64\n",
      " 22  momentum_12    77788 non-null  float64\n",
      " 23  momentum_3_12  77788 non-null  float64\n",
      " 24  year           77788 non-null  int64  \n",
      " 25  month          77788 non-null  int64  \n",
      " 26  target         77788 non-null  float64\n",
      "dtypes: float64(24), int64(2), object(1)\n",
      "memory usage: 16.4+ MB\n"
     ]
    }
   ],
   "source": [
    "data.sort_index().info(null_counts=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sector Breakdown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:27.081741Z",
     "start_time": "2021-02-23T20:42:26.879546Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = data.reset_index().groupby('sector').ticker.nunique().sort_values().plot.barh(title='Sector Breakdown')\n",
    "ax.set_ylabel('')\n",
    "ax.set_xlabel('# Tickers')\n",
    "sns.despine()\n",
    "plt.tight_layout();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Store data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:30.009591Z",
     "start_time": "2021-02-23T20:42:29.969076Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "with pd.HDFStore('data.h5') as store:\n",
    "    store.put('us/equities/monthly', data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate mutual information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:30.620609Z",
     "start_time": "2021-02-23T20:42:30.607499Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "X = data.drop('target', axis=1)\n",
    "X.sector = pd.factorize(X.sector)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:45.996017Z",
     "start_time": "2021-02-23T20:42:30.768783Z"
    }
   },
   "outputs": [],
   "source": [
    "mi = mutual_info_regression(X=X, y=data.target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:46.001333Z",
     "start_time": "2021-02-23T20:42:45.997201Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "natr             0.111798\n",
       "return_12m       0.060056\n",
       "return_6m        0.054047\n",
       "year             0.049061\n",
       "return_3m        0.047197\n",
       "momentum_3_12    0.039969\n",
       "momentum_3_6     0.037950\n",
       "bb_up            0.035975\n",
       "momentum_12      0.035818\n",
       "return_1m        0.034792\n",
       "dtype: float64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mi_reg = pd.Series(mi, index=X.columns)\n",
    "mi_reg.nlargest(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:56.372539Z",
     "start_time": "2021-02-23T20:42:46.002764Z"
    }
   },
   "outputs": [],
   "source": [
    "mi = mutual_info_classif(X=X, y=(data.target>0).astype(int))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:56.378061Z",
     "start_time": "2021-02-23T20:42:56.373817Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year          0.011498\n",
       "month         0.006240\n",
       "atr           0.005068\n",
       "return_6m     0.004963\n",
       "rsi           0.004232\n",
       "return_12m    0.003727\n",
       "sector        0.003231\n",
       "RMW           0.002064\n",
       "natr          0.001931\n",
       "return_1m     0.001754\n",
       "dtype: float64"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mi_class = pd.Series(mi, index=X.columns)\n",
    "mi_class.nlargest(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:56.402392Z",
     "start_time": "2021-02-23T20:42:56.379294Z"
    }
   },
   "outputs": [],
   "source": [
    "mi = mi_reg.to_frame('Regression').join(mi_class.to_frame('Classification'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:56.410287Z",
     "start_time": "2021-02-23T20:42:56.403851Z"
    }
   },
   "outputs": [],
   "source": [
    "mi.index = [' '.join(c.upper().split('_')) for c in mi.index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-02-23T20:42:56.986502Z",
     "start_time": "2021-02-23T20:42:56.411402Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(ncols=2, figsize=(12, 4))\n",
    "for i, t in enumerate(['Regression', 'Classification']):\n",
    "    mi[t].nlargest(20).sort_values().plot.barh(title=t, ax=axes[i])\n",
    "    axes[i].set_xlabel('Mutual Information')\n",
    "fig.suptitle('Mutual Information', fontsize=14)\n",
    "sns.despine()\n",
    "fig.tight_layout()\n",
    "fig.subplots_adjust(top=.9)"
   ]
  }
 ],
 "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": "3.8.8"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "230.355px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}