2_prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "395bb63e-48ca-4f97-a88b-df4fd0a19615",
   "metadata": {},
   "source": [
    "# Introduction\n",
    "\n",
    "The main purpose of this repository is to show how machine learning can be used to *understand what is happening and why*, rather than just for prediction. However, the data setup is often the same for both, so I may as well do both.\n",
    "\n",
    "Here, I will attempt to predict a company's value (market cap). To do so, I use Optuna to tune an xgboost regressor. I then train the winning model and do a bit of performance analysis, finishing with a quick list of \"todos\".\n",
    "\n",
    "In the next notebook I'll demonstrate how to use ML to teach us from the data instead."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23807f58-f825-4866-9e90-bf795f2f8d40",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "Disclaimer: I am not an investment professional. None of my work within or related to this notebook should be considered investment advice. It is not."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "93cba46a-a007-47b6-818f-56dbf8b2612d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import duckdb\n",
    "import os\n",
    "import pandas as pd\n",
    "pd.set_option('display.float_format', lambda x: '%.3f' % x)\n",
    "import numpy as np\n",
    "from pathlib import Path\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('dark_background')\n",
    "from datetime import datetime\n",
    "\n",
    "data_path = Path(os.getenv('DATA_HOME'))\n",
    "duckdb_path = data_path.joinpath('analytics_demo/sharadar.duckdb')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40026b47-a208-400a-bb8f-369dbdf86754",
   "metadata": {},
   "source": [
    "# Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "1a267fce-af4f-4a05-ac48-1ebe7be2258c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e8a5a8f967514b588d30534a8a19284b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "FloatProgress(value=0.0, layout=Layout(width='auto'), style=ProgressStyle(bar_color='black'))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fct_daily_targ: (1764497, 14)\n"
     ]
    },
    {
     "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>key_daily</th>\n",
       "      <th>ticker</th>\n",
       "      <th>date</th>\n",
       "      <th>volume</th>\n",
       "      <th>closeadj</th>\n",
       "      <th>mcap_daily</th>\n",
       "      <th>ev_daily</th>\n",
       "      <th>price_chg_1y</th>\n",
       "      <th>excess_price_chg_1y</th>\n",
       "      <th>key_dim_fin</th>\n",
       "      <th>mcap_rank</th>\n",
       "      <th>mcap_perc_rank</th>\n",
       "      <th>mcap_rank_rev</th>\n",
       "      <th>mcap_perc_rank_rev</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TRTN_2017-07-06</td>\n",
       "      <td>TRTN</td>\n",
       "      <td>2017-07-06</td>\n",
       "      <td>290500.000</td>\n",
       "      <td>24.000</td>\n",
       "      <td>2467400000.000</td>\n",
       "      <td>8759100000.000</td>\n",
       "      <td>0.058</td>\n",
       "      <td>-0.043</td>\n",
       "      <td>TRTN_ART_2017-03-31_2017-05-15</td>\n",
       "      <td>1510</td>\n",
       "      <td>0.292</td>\n",
       "      <td>3664</td>\n",
       "      <td>0.708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>UST1_1999-01-06</td>\n",
       "      <td>UST1</td>\n",
       "      <td>1999-01-06</td>\n",
       "      <td>404200.000</td>\n",
       "      <td>19.720</td>\n",
       "      <td>6392400000.000</td>\n",
       "      <td>6371300000.000</td>\n",
       "      <td>-0.131</td>\n",
       "      <td>-0.181</td>\n",
       "      <td>UST1_ART_1998-09-30_1998-11-10</td>\n",
       "      <td>357</td>\n",
       "      <td>0.050</td>\n",
       "      <td>6767</td>\n",
       "      <td>0.950</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         key_daily ticker       date     volume  closeadj     mcap_daily  \\\n",
       "0  TRTN_2017-07-06   TRTN 2017-07-06 290500.000    24.000 2467400000.000   \n",
       "1  UST1_1999-01-06   UST1 1999-01-06 404200.000    19.720 6392400000.000   \n",
       "\n",
       "        ev_daily  price_chg_1y  excess_price_chg_1y  \\\n",
       "0 8759100000.000         0.058               -0.043   \n",
       "1 6371300000.000        -0.131               -0.181   \n",
       "\n",
       "                      key_dim_fin  mcap_rank  mcap_perc_rank  mcap_rank_rev  \\\n",
       "0  TRTN_ART_2017-03-31_2017-05-15       1510           0.292           3664   \n",
       "1  UST1_ART_1998-09-30_1998-11-10        357           0.050           6767   \n",
       "\n",
       "   mcap_perc_rank_rev  \n",
       "0               0.708  \n",
       "1               0.950  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df_mcap_reg: (1131433, 877)\n"
     ]
    },
    {
     "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>key_daily</th>\n",
       "      <th>date</th>\n",
       "      <th>mcap_daily_scaled_log</th>\n",
       "      <th>accoci_art_scaled</th>\n",
       "      <th>assets_art_scaled</th>\n",
       "      <th>assetsavg_art_scaled</th>\n",
       "      <th>assetsc_art_scaled</th>\n",
       "      <th>assetsnc_art_scaled</th>\n",
       "      <th>capex_art_scaled</th>\n",
       "      <th>cashneq_art_scaled</th>\n",
       "      <th>...</th>\n",
       "      <th>liabilitiesc_over_fcf_art_chg</th>\n",
       "      <th>payables_over_assetsavg_art_chg</th>\n",
       "      <th>payables_over_revenue_art_chg</th>\n",
       "      <th>payables_over_fcf_art_chg</th>\n",
       "      <th>ncfx_over_assetsavg_art_chg</th>\n",
       "      <th>ncfx_over_revenue_art_chg</th>\n",
       "      <th>ncfx_over_fcf_art_chg</th>\n",
       "      <th>deferredrev_over_assetsavg_art_chg</th>\n",
       "      <th>deferredrev_over_revenue_art_chg</th>\n",
       "      <th>deferredrev_over_fcf_art_chg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CLGRF_2015-03-04</td>\n",
       "      <td>2015-03-04</td>\n",
       "      <td>9.203</td>\n",
       "      <td>0.209</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.999</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.545</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.016</td>\n",
       "      <td>-0.084</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.000</td>\n",
       "      <td>-0.009</td>\n",
       "      <td>-0.039</td>\n",
       "      <td>0.148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>WKSP_2020-11-04</td>\n",
       "      <td>2020-11-04</td>\n",
       "      <td>5.968</td>\n",
       "      <td>0.082</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.158</td>\n",
       "      <td>0.597</td>\n",
       "      <td>0.357</td>\n",
       "      <td>0.399</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.030</td>\n",
       "      <td>-0.073</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 877 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          key_daily       date  mcap_daily_scaled_log  accoci_art_scaled  \\\n",
       "0  CLGRF_2015-03-04 2015-03-04                  9.203              0.209   \n",
       "1   WKSP_2020-11-04 2020-11-04                  5.968              0.082   \n",
       "\n",
       "   assets_art_scaled  assetsavg_art_scaled  assetsc_art_scaled  \\\n",
       "0              0.000                 0.000               0.000   \n",
       "1              0.000                 0.000               0.000   \n",
       "\n",
       "   assetsnc_art_scaled  capex_art_scaled  cashneq_art_scaled  ...  \\\n",
       "0                0.000             0.999               0.000  ...   \n",
       "1                0.000             1.000               0.000  ...   \n",
       "\n",
       "   liabilitiesc_over_fcf_art_chg  payables_over_assetsavg_art_chg  \\\n",
       "0                         -1.545                            0.007   \n",
       "1                          1.158                            0.597   \n",
       "\n",
       "   payables_over_revenue_art_chg  payables_over_fcf_art_chg  \\\n",
       "0                          0.016                     -0.084   \n",
       "1                          0.357                      0.399   \n",
       "\n",
       "   ncfx_over_assetsavg_art_chg  ncfx_over_revenue_art_chg  \\\n",
       "0                        0.000                      0.000   \n",
       "1                        0.031                      0.030   \n",
       "\n",
       "   ncfx_over_fcf_art_chg  deferredrev_over_assetsavg_art_chg  \\\n",
       "0                 -0.000                              -0.009   \n",
       "1                 -0.073                               0.000   \n",
       "\n",
       "   deferredrev_over_revenue_art_chg  deferredrev_over_fcf_art_chg  \n",
       "0                            -0.039                         0.148  \n",
       "1                             0.000                        -0.000  \n",
       "\n",
       "[2 rows x 877 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "con = duckdb.connect(str(duckdb_path))\n",
    "df_fct_daily_targ = con.execute('from fct_daily_targ where mcap_daily > 0').fetch_df()\n",
    "df_mcap_reg = con.execute('from mcap_reg').fetch_df()\n",
    "con.close()\n",
    "\n",
    "print('fct_daily_targ:', df_fct_daily_targ.shape)\n",
    "display(df_fct_daily_targ.head(2))\n",
    "print('df_mcap_reg:', df_mcap_reg.shape)\n",
    "display(df_mcap_reg.head(2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34320f23-738c-4b04-bbbf-1cc18358776e",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "# EDA &  Prep\n",
    "\n",
    "I'm going to use an expanding window for tuning, training through a given year then predicting a few following years' worth of market cap values.\n",
    "\n",
    "![Time Series ML Folds](./images/time_series_ml_folds.png)\n",
    "\n",
    "Tasks:\n",
    "- Impute +/- inf.\n",
    "- Drop features with too many nulls.\n",
    "- Find a \"starting year\" for the first training window (\"TW1\").\n",
    "- To speed up tuning, use TW1 to build a correlation matrix and drop highly-correlated features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "17499618-0c64-4a35-8b62-64453eb9b321",
   "metadata": {},
   "outputs": [],
   "source": [
    "# I'll be using XGBoost, which can handle nulls. Since this is a demo, I'll simply set inf to null.\n",
    "df_mcap_reg = df_mcap_reg.replace([-np.inf, np.inf], np.nan)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ca395a62-3623-4509-8d34-f3103e880f43",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols_feat = df_mcap_reg.columns.drop(['key_daily', 'date', 'mcap_daily_scaled_log']).tolist()\n",
    "\n",
    "df_frac_nullity = df_mcap_reg[cols_feat].isnull().mean()\n",
    "df_frac_nullity.hist(bins=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4bc2e4b1-815c-4aef-ad14-95ffe5f1d822",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 feature columns dropped for having too many nulls.\n"
     ]
    }
   ],
   "source": [
    "# Arbitrary, but hey.\n",
    "frac_too_many_nulls = .4\n",
    "\n",
    "cols_feat_drop_for_nullity = set(df_frac_nullity[df_frac_nullity > frac_too_many_nulls].index)\n",
    "print(\"{} feature columns dropped for having too many nulls.\".format(str(len(cols_feat_drop_for_nullity))))\n",
    "cols_feat = list(set(cols_feat).difference(cols_feat_drop_for_nullity))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3e046d55-dffb-471f-b612-6f2dfd3c0b52",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='date'>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "date\n",
       "1998       56\n",
       "1999      880\n",
       "2000     1077\n",
       "2001    13858\n",
       "2002    39286\n",
       "2003    51210\n",
       "2004    53351\n",
       "2005    53748\n",
       "2006    52823\n",
       "2007    52057\n",
       "2008    51048\n",
       "Name: count, dtype: int64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute daily and yearly row counts\n",
    "display(df_mcap_reg.date.value_counts().sort_index().plot())\n",
    "display(df_mcap_reg.date.dt.year.value_counts().sort_index().head(11))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6e6ab8d-870b-4d1a-b23d-670a886a2fa7",
   "metadata": {},
   "source": [
    "Let's extend TW1 through 2007, which will give it 5 full years. Below I create train/predict folds in a way that gives prediction coverage from 2008 to the present year without the need to train one model for each year."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "87670fc9-c843-4552-bf3c-0a41a6ec16b6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[..., 2007] --> [2008, 2009, 2010, 2011, 2012]\n",
      "[..., 2012] --> [2013, 2014, 2015, 2016, 2017]\n",
      "[..., 2017] --> [2018, 2019, 2020, 2021, 2022]\n",
      "[..., 2022] --> [2023, 2024, 2025, 2026, 2027]\n"
     ]
    }
   ],
   "source": [
    "# Store training window partition\n",
    "first_window_final_year = 2007\n",
    "years_btw_windows = 5\n",
    "\n",
    "current_year = datetime.utcnow().year\n",
    "train_window_limits = list(np.arange(first_window_final_year, current_year, years_btw_windows))\n",
    "pred_window_lists = [list(range(x + 1, x + years_btw_windows + 1)) for x in train_window_limits]\n",
    "\n",
    "for ty, py in zip(train_window_limits, pred_window_lists):\n",
    "    print('[..., {}]'.format(ty), '-->', py)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4795f2a9-1d06-4a92-b1b1-82269a6f7c6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Will drop 432 of 874 features.\n"
     ]
    }
   ],
   "source": [
    "#### For variable pairs with this absolute correlation or greater, keep only one.\n",
    "max_corr = 0.95\n",
    "\n",
    "# Sample within the first train window then compute correlation matrix.\n",
    "tw1_limit = min(train_window_limits)\n",
    "X_corr = df_mcap_reg.loc[df_mcap_reg.date.dt.year <= tw1_limit, cols_feat].sample(n=10000).corr()\n",
    "\n",
    "# Convert to upper triangle then find list of columns with strong correlation.\n",
    "X_corr = X_corr.where(np.triu(np.ones(X_corr.shape), k=1).astype(bool))\n",
    "correlated_feat_drop = [c for c in X_corr.columns if any(X_corr[c].abs() > max_corr)]\n",
    "print(\"Will drop {} of {} features.\".format(str(len(correlated_feat_drop)), str(len(cols_feat))))\n",
    "\n",
    "cols_feat_keep_low_corr = list(set(X_corr.columns).difference(set(correlated_feat_drop)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "5bd0034f-cc55-4239-a263-2b0264783c58",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cudf\n",
    "import cupy\n",
    "from itertools import product\n",
    "import optuna\n",
    "from sklearn.metrics import mean_absolute_percentage_error as mape_score\n",
    "import xgboost\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore', category=FutureWarning)\n",
    "\n",
    "import joblib\n",
    "import os\n",
    "from pathlib import Path\n",
    "study_filepath = data_path.joinpath(\"analytics_demo\").joinpath('supervised_ml_study.pkl')\n",
    "best_model_filepath = data_path.joinpath(\"analytics_demo\").joinpath('reg_xgb_best.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c5bffb5-abb5-4830-8331-b0a474e93211",
   "metadata": {},
   "source": [
    "# Tune ML\n",
    "\n",
    "For hyperparameter tuning, rather than using just a grid *or* an optimizer, I like to force a basic grid search *then* run an optimizer. Otherwise they seem prone to getting stuck at local optima. Optuna's `enqueue_trial()` allows for just this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c0b9b047-1467-487c-bba7-8318c4d2d078",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1131433, 445)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cudf_reg = cudf.from_pandas(df_mcap_reg[['key_daily', 'date', 'mcap_daily_scaled_log'] + cols_feat_keep_low_corr])\n",
    "cudf_reg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4df70144-6468-42c1-b514-151283700f6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create boolean masks, which I call selectors, for train and prediction windows.\n",
    "train_window_selectors = [cudf_reg.date.dt.year <= cutoff_yr for cutoff_yr in train_window_limits]\n",
    "prediction_window_selectors = [cudf_reg.date.dt.year.isin(pwl) for pwl in pred_window_lists]\n",
    "window_selectors = list(zip(train_window_selectors, prediction_window_selectors))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "6d7f1532-7d0a-4a58-ad17-4adab84a31d4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "96 models in grid.\n"
     ]
    },
    {
     "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>n_estimators</th>\n",
       "      <th>learning_rate</th>\n",
       "      <th>max_depth</th>\n",
       "      <th>colsample_bytree</th>\n",
       "      <th>subsample</th>\n",
       "      <th>min_child_weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>25</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.100</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>25</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.100</td>\n",
       "      <td>500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>25</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.100</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.500</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>25</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.500</td>\n",
       "      <td>500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   n_estimators  learning_rate  max_depth  colsample_bytree  subsample  \\\n",
       "0            25          0.001          3             0.100      0.100   \n",
       "1            25          0.001          3             0.100      0.100   \n",
       "2            25          0.001          3             0.100      0.100   \n",
       "3            25          0.001          3             0.100      0.500   \n",
       "4            25          0.001          3             0.100      0.500   \n",
       "\n",
       "   min_child_weight  \n",
       "0                50  \n",
       "1               500  \n",
       "2              5000  \n",
       "3                50  \n",
       "4               500  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_labels = ['n_estimators', 'learning_rate', 'max_depth', 'colsample_bytree', 'subsample', 'min_child_weight']\n",
    "\n",
    "grid_n_estimators = [25, 250]\n",
    "grid_learning_rate = [0.001, 0.01]\n",
    "grid_max_depth = [3, 10]\n",
    "grid_colsample_bytree = [0.1, 0.5]\n",
    "grid_subsample = grid_colsample_bytree\n",
    "grid_min_child_weight = [50, 500, 5000]\n",
    "\n",
    "grid = product(grid_n_estimators, grid_learning_rate, grid_max_depth, grid_colsample_bytree, grid_subsample, grid_min_child_weight)\n",
    "grid = [{param_labels[i]: g[i] for i in range(len(param_labels))} for g in grid]\n",
    "print(len(grid), 'models in grid.')\n",
    "pd.DataFrame(grid).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "44e19835-e450-4062-b179-6f42740e8a3b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[I 2024-05-21 02:59:44,477] A new study created in memory with name: no-name-5911cc79-93b1-40bf-bbfb-d707330cbbce\n"
     ]
    }
   ],
   "source": [
    "study = optuna.create_study(direction='minimize')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "41319739-65bd-4633-a80e-7c89204c2b44",
   "metadata": {},
   "outputs": [],
   "source": [
    "for _t in grid:\n",
    "    study.enqueue_trial(_t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "5179563f-8d61-4922-9eb8-f2a14f804677",
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_trial(trial):\n",
    "    global cudf_reg\n",
    "    cudf_reg['pred'] = cupy.nan\n",
    "    \n",
    "    model_params = {\n",
    "        'device': 'cuda',  # For xgboost>=2.0.0\n",
    "        'tree_method': 'hist',  # For xgboost>=2.0.0\n",
    "        'objective': 'reg:squarederror',\n",
    "        'n_estimators': trial.suggest_int('n_estimators', 25, 500),\n",
    "        'learning_rate': trial.suggest_float('learning_rate', 0.00001, 10),\n",
    "        'max_depth': trial.suggest_int('max_depth', 1, 15),\n",
    "        'colsample_bytree': trial.suggest_float('colsample_bytree', 0.01,1.0),\n",
    "        'subsample': trial.suggest_float('subsample', 0.01, 1.0),\n",
    "        'min_child_weight': trial.suggest_int('min_child_weight', 1, 100000)  # leaf sample cnt\n",
    "    }\n",
    "   \n",
    "    for sel_train, sel_pred in window_selectors:\n",
    "        reg = xgboost.XGBRegressor(**model_params)\n",
    "        reg.fit(X=cudf_reg.loc[sel_train, cols_feat_keep_low_corr], y=cudf_reg.loc[sel_train, 'mcap_daily_scaled_log'])\n",
    "        cudf_reg.loc[sel_pred, 'pred'] = reg.predict(X=cudf_reg.loc[sel_pred, cols_feat_keep_low_corr])\n",
    "\n",
    "    df_score = cudf_reg.loc[~cudf_reg.pred.isnull(), ['mcap_daily_scaled_log', 'pred']].to_pandas()\n",
    "    return mape_score(df_score.mcap_daily_scaled_log, df_score.pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e33d154e-6d3d-4340-b0b5-0cab6c349b5a",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "study.optimize(run_trial, 500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "2e0651a9-a5ac-4856-a7ac-4a3eab69a39d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read or write study to disk\n",
    "# study = joblib.load(study_filepath)\n",
    "joblib.dump(study, study_filepath);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f49af78c-1e74-4c69-b6dd-c17152ab016e",
   "metadata": {},
   "source": [
    "# Tuning results and analysis\n",
    "\n",
    "Below we see what I generally hope to see within the top trial runs: solid trends in most of the hyperparameters. That means we've got a region in the hyperparam space which repeatedly performs well.\n",
    "\n",
    "There's more you'd want to look at here. For instance, this \"high performance region\" in the parameter space actually may perform unreliably. That is, it may generate the very best and worst trial runs, yielding mediocre average performance. I like to use k-means clustering to study results like this, especialy when doing multi-target hyperparam optimization.\n",
    "\n",
    "But for now, we'll just use the top parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "620cabdf-7171-43ca-aab2-ac52fab62197",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_c31d0_row0_col0, #T_c31d0_row5_col4, #T_c31d0_row6_col6, #T_c31d0_row8_col3, #T_c31d0_row11_col2, #T_c31d0_row13_col1, #T_c31d0_row15_col3, #T_c31d0_row16_col3, #T_c31d0_row16_col5 {\n",
       "  background-color: #3b4cc0;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row0_col1 {\n",
       "  background-color: #b3cdfb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row0_col2 {\n",
       "  background-color: #4c66d6;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row0_col3, #T_c31d0_row1_col3, #T_c31d0_row2_col3, #T_c31d0_row3_col3, #T_c31d0_row7_col3, #T_c31d0_row9_col3, #T_c31d0_row10_col3, #T_c31d0_row11_col3, #T_c31d0_row12_col3, #T_c31d0_row13_col3, #T_c31d0_row14_col3, #T_c31d0_row17_col3, #T_c31d0_row18_col3, #T_c31d0_row19_col3 {\n",
       "  background-color: #dddcdc;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row0_col4, #T_c31d0_row4_col5 {\n",
       "  background-color: #6384eb;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row0_col5, #T_c31d0_row19_col5 {\n",
       "  background-color: #9fbfff;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row0_col6 {\n",
       "  background-color: #7a9df8;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row1_col0 {\n",
       "  background-color: #6788ee;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row1_col1, #T_c31d0_row16_col4 {\n",
       "  background-color: #d3dbe7;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row1_col2 {\n",
       "  background-color: #e67259;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row1_col4, #T_c31d0_row4_col4, #T_c31d0_row6_col4 {\n",
       "  background-color: #4358cb;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row1_col5, #T_c31d0_row7_col6, #T_c31d0_row19_col4 {\n",
       "  background-color: #afcafc;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row1_col6 {\n",
       "  background-color: #aec9fc;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row2_col0, #T_c31d0_row10_col6 {\n",
       "  background-color: #688aef;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row2_col1 {\n",
       "  background-color: #bcd2f7;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row2_col2 {\n",
       "  background-color: #5d7ce6;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row2_col4 {\n",
       "  background-color: #b2ccfb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row2_col5, #T_c31d0_row3_col5, #T_c31d0_row15_col5 {\n",
       "  background-color: #ecd3c5;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row2_col6, #T_c31d0_row3_col1 {\n",
       "  background-color: #b7cff9;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row3_col0, #T_c31d0_row12_col5 {\n",
       "  background-color: #bed2f6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row3_col2, #T_c31d0_row6_col2 {\n",
       "  background-color: #e46e56;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row3_col4 {\n",
       "  background-color: #5673e0;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row3_col6 {\n",
       "  background-color: #a9c6fd;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row4_col0, #T_c31d0_row18_col4 {\n",
       "  background-color: #c6d6f1;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row4_col1 {\n",
       "  background-color: #3f53c6;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row4_col2, #T_c31d0_row12_col0 {\n",
       "  background-color: #f7a688;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row4_col3, #T_c31d0_row5_col3, #T_c31d0_row6_col3, #T_c31d0_row6_col5, #T_c31d0_row8_col4, #T_c31d0_row9_col1, #T_c31d0_row9_col5, #T_c31d0_row9_col6, #T_c31d0_row16_col2, #T_c31d0_row19_col0 {\n",
       "  background-color: #b40426;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row4_col6, #T_c31d0_row14_col1 {\n",
       "  background-color: #7396f5;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row5_col0 {\n",
       "  background-color: #ccd9ed;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row5_col1 {\n",
       "  background-color: #7b9ff9;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row5_col2 {\n",
       "  background-color: #f7af91;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row5_col5, #T_c31d0_row8_col5, #T_c31d0_row10_col5, #T_c31d0_row13_col5, #T_c31d0_row18_col5 {\n",
       "  background-color: #7295f4;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row5_col6 {\n",
       "  background-color: #88abfd;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row6_col0 {\n",
       "  background-color: #cedaeb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row6_col1, #T_c31d0_row7_col4 {\n",
       "  background-color: #5a78e4;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row7_col0 {\n",
       "  background-color: #e5d8d1;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row7_col1, #T_c31d0_row13_col6 {\n",
       "  background-color: #b5cdfa;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row7_col2 {\n",
       "  background-color: #f7b599;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row7_col5 {\n",
       "  background-color: #d7dce3;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row8_col0 {\n",
       "  background-color: #f3c7b1;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row8_col1, #T_c31d0_row13_col0 {\n",
       "  background-color: #ed8366;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row8_col2 {\n",
       "  background-color: #536edd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row8_col6 {\n",
       "  background-color: #f59f80;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row9_col0 {\n",
       "  background-color: #f6bfa6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row9_col2, #T_c31d0_row15_col2 {\n",
       "  background-color: #5470de;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row9_col4 {\n",
       "  background-color: #c4d5f3;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row10_col0 {\n",
       "  background-color: #f7ba9f;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row10_col1 {\n",
       "  background-color: #9ebeff;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row10_col2, #T_c31d0_row12_col6, #T_c31d0_row17_col5 {\n",
       "  background-color: #f2c9b4;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row10_col4, #T_c31d0_row13_col4 {\n",
       "  background-color: #cad8ef;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row11_col0, #T_c31d0_row16_col6 {\n",
       "  background-color: #f7b89c;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row11_col1 {\n",
       "  background-color: #f29274;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row11_col4 {\n",
       "  background-color: #a2c1ff;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row11_col5 {\n",
       "  background-color: #f59d7e;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row11_col6 {\n",
       "  background-color: #d55042;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row12_col1 {\n",
       "  background-color: #d5dbe5;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row12_col2 {\n",
       "  background-color: #4a63d3;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row12_col4 {\n",
       "  background-color: #e1dad6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row13_col2 {\n",
       "  background-color: #ec7f63;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row14_col0 {\n",
       "  background-color: #e8765c;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row14_col2 {\n",
       "  background-color: #d75445;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row14_col4 {\n",
       "  background-color: #4055c8;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row14_col5 {\n",
       "  background-color: #5572df;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row14_col6 {\n",
       "  background-color: #7ea1fa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row15_col0 {\n",
       "  background-color: #e57058;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row15_col1 {\n",
       "  background-color: #f5c0a7;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row15_col4 {\n",
       "  background-color: #c12b30;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row15_col6 {\n",
       "  background-color: #da5a49;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row16_col0 {\n",
       "  background-color: #d1493f;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row16_col1 {\n",
       "  background-color: #84a7fc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row17_col0 {\n",
       "  background-color: #d0473d;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row17_col1 {\n",
       "  background-color: #c9d7f0;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row17_col2 {\n",
       "  background-color: #6485ec;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row17_col4 {\n",
       "  background-color: #ebd3c6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row17_col6 {\n",
       "  background-color: #f4987a;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row18_col0 {\n",
       "  background-color: #cb3e38;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row18_col1, #T_c31d0_row19_col6 {\n",
       "  background-color: #bad0f8;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row18_col2 {\n",
       "  background-color: #c53334;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_c31d0_row18_col6 {\n",
       "  background-color: #b1cbfc;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row19_col1 {\n",
       "  background-color: #aac7fd;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_c31d0_row19_col2 {\n",
       "  background-color: #ef886b;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_c31d0\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_c31d0_level0_col0\" class=\"col_heading level0 col0\" >mape</th>\n",
       "      <th id=\"T_c31d0_level0_col1\" class=\"col_heading level0 col1\" >params_colsample_bytree</th>\n",
       "      <th id=\"T_c31d0_level0_col2\" class=\"col_heading level0 col2\" >params_learning_rate</th>\n",
       "      <th id=\"T_c31d0_level0_col3\" class=\"col_heading level0 col3\" >params_max_depth</th>\n",
       "      <th id=\"T_c31d0_level0_col4\" class=\"col_heading level0 col4\" >params_min_child_weight</th>\n",
       "      <th id=\"T_c31d0_level0_col5\" class=\"col_heading level0 col5\" >params_n_estimators</th>\n",
       "      <th id=\"T_c31d0_level0_col6\" class=\"col_heading level0 col6\" >params_subsample</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row0\" class=\"row_heading level0 row0\" >392</th>\n",
       "      <td id=\"T_c31d0_row0_col0\" class=\"data row0 col0\" >0.09614</td>\n",
       "      <td id=\"T_c31d0_row0_col1\" class=\"data row0 col1\" >0.82337</td>\n",
       "      <td id=\"T_c31d0_row0_col2\" class=\"data row0 col2\" >0.15689</td>\n",
       "      <td id=\"T_c31d0_row0_col3\" class=\"data row0 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row0_col4\" class=\"data row0 col4\" >3813</td>\n",
       "      <td id=\"T_c31d0_row0_col5\" class=\"data row0 col5\" >484</td>\n",
       "      <td id=\"T_c31d0_row0_col6\" class=\"data row0 col6\" >0.85679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row1\" class=\"row_heading level0 row1\" >408</th>\n",
       "      <td id=\"T_c31d0_row1_col0\" class=\"data row1 col0\" >0.09659</td>\n",
       "      <td id=\"T_c31d0_row1_col1\" class=\"data row1 col1\" >0.84106</td>\n",
       "      <td id=\"T_c31d0_row1_col2\" class=\"data row1 col2\" >0.35667</td>\n",
       "      <td id=\"T_c31d0_row1_col3\" class=\"data row1 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row1_col4\" class=\"data row1 col4\" >2370</td>\n",
       "      <td id=\"T_c31d0_row1_col5\" class=\"data row1 col5\" >485</td>\n",
       "      <td id=\"T_c31d0_row1_col6\" class=\"data row1 col6\" >0.87905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row2\" class=\"row_heading level0 row2\" >391</th>\n",
       "      <td id=\"T_c31d0_row2_col0\" class=\"data row2 col0\" >0.09660</td>\n",
       "      <td id=\"T_c31d0_row2_col1\" class=\"data row2 col1\" >0.82820</td>\n",
       "      <td id=\"T_c31d0_row2_col2\" class=\"data row2 col2\" >0.16990</td>\n",
       "      <td id=\"T_c31d0_row2_col3\" class=\"data row2 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row2_col4\" class=\"data row2 col4\" >7082</td>\n",
       "      <td id=\"T_c31d0_row2_col5\" class=\"data row2 col5\" >490</td>\n",
       "      <td id=\"T_c31d0_row2_col6\" class=\"data row2 col6\" >0.88313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row3\" class=\"row_heading level0 row3\" >395</th>\n",
       "      <td id=\"T_c31d0_row3_col0\" class=\"data row3 col0\" >0.09737</td>\n",
       "      <td id=\"T_c31d0_row3_col1\" class=\"data row3 col1\" >0.82501</td>\n",
       "      <td id=\"T_c31d0_row3_col2\" class=\"data row3 col2\" >0.35866</td>\n",
       "      <td id=\"T_c31d0_row3_col3\" class=\"data row3 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row3_col4\" class=\"data row3 col4\" >3300</td>\n",
       "      <td id=\"T_c31d0_row3_col5\" class=\"data row3 col5\" >490</td>\n",
       "      <td id=\"T_c31d0_row3_col6\" class=\"data row3 col6\" >0.87656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row4\" class=\"row_heading level0 row4\" >421</th>\n",
       "      <td id=\"T_c31d0_row4_col0\" class=\"data row4 col0\" >0.09746</td>\n",
       "      <td id=\"T_c31d0_row4_col1\" class=\"data row4 col1\" >0.76261</td>\n",
       "      <td id=\"T_c31d0_row4_col2\" class=\"data row4 col2\" >0.32569</td>\n",
       "      <td id=\"T_c31d0_row4_col3\" class=\"data row4 col3\" >12</td>\n",
       "      <td id=\"T_c31d0_row4_col4\" class=\"data row4 col4\" >2361</td>\n",
       "      <td id=\"T_c31d0_row4_col5\" class=\"data row4 col5\" >480</td>\n",
       "      <td id=\"T_c31d0_row4_col6\" class=\"data row4 col6\" >0.85361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row5\" class=\"row_heading level0 row5\" >420</th>\n",
       "      <td id=\"T_c31d0_row5_col0\" class=\"data row5 col0\" >0.09751</td>\n",
       "      <td id=\"T_c31d0_row5_col1\" class=\"data row5 col1\" >0.79513</td>\n",
       "      <td id=\"T_c31d0_row5_col2\" class=\"data row5 col2\" >0.31936</td>\n",
       "      <td id=\"T_c31d0_row5_col3\" class=\"data row5 col3\" >12</td>\n",
       "      <td id=\"T_c31d0_row5_col4\" class=\"data row5 col4\" >1968</td>\n",
       "      <td id=\"T_c31d0_row5_col5\" class=\"data row5 col5\" >481</td>\n",
       "      <td id=\"T_c31d0_row5_col6\" class=\"data row5 col6\" >0.86268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row6\" class=\"row_heading level0 row6\" >424</th>\n",
       "      <td id=\"T_c31d0_row6_col0\" class=\"data row6 col0\" >0.09754</td>\n",
       "      <td id=\"T_c31d0_row6_col1\" class=\"data row6 col1\" >0.77790</td>\n",
       "      <td id=\"T_c31d0_row6_col2\" class=\"data row6 col2\" >0.35903</td>\n",
       "      <td id=\"T_c31d0_row6_col3\" class=\"data row6 col3\" >12</td>\n",
       "      <td id=\"T_c31d0_row6_col4\" class=\"data row6 col4\" >2380</td>\n",
       "      <td id=\"T_c31d0_row6_col5\" class=\"data row6 col5\" >500</td>\n",
       "      <td id=\"T_c31d0_row6_col6\" class=\"data row6 col6\" >0.82705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row7\" class=\"row_heading level0 row7\" >393</th>\n",
       "      <td id=\"T_c31d0_row7_col0\" class=\"data row7 col0\" >0.09781</td>\n",
       "      <td id=\"T_c31d0_row7_col1\" class=\"data row7 col1\" >0.82424</td>\n",
       "      <td id=\"T_c31d0_row7_col2\" class=\"data row7 col2\" >0.31446</td>\n",
       "      <td id=\"T_c31d0_row7_col3\" class=\"data row7 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row7_col4\" class=\"data row7 col4\" >3470</td>\n",
       "      <td id=\"T_c31d0_row7_col5\" class=\"data row7 col5\" >488</td>\n",
       "      <td id=\"T_c31d0_row7_col6\" class=\"data row7 col6\" >0.87934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row8\" class=\"row_heading level0 row8\" >349</th>\n",
       "      <td id=\"T_c31d0_row8_col0\" class=\"data row8 col0\" >0.09807</td>\n",
       "      <td id=\"T_c31d0_row8_col1\" class=\"data row8 col1\" >0.90137</td>\n",
       "      <td id=\"T_c31d0_row8_col2\" class=\"data row8 col2\" >0.16160</td>\n",
       "      <td id=\"T_c31d0_row8_col3\" class=\"data row8 col3\" >10</td>\n",
       "      <td id=\"T_c31d0_row8_col4\" class=\"data row8 col4\" >16218</td>\n",
       "      <td id=\"T_c31d0_row8_col5\" class=\"data row8 col5\" >481</td>\n",
       "      <td id=\"T_c31d0_row8_col6\" class=\"data row8 col6\" >0.93742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row9\" class=\"row_heading level0 row9\" >384</th>\n",
       "      <td id=\"T_c31d0_row9_col0\" class=\"data row9 col0\" >0.09816</td>\n",
       "      <td id=\"T_c31d0_row9_col1\" class=\"data row9 col1\" >0.93592</td>\n",
       "      <td id=\"T_c31d0_row9_col2\" class=\"data row9 col2\" >0.16281</td>\n",
       "      <td id=\"T_c31d0_row9_col3\" class=\"data row9 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row9_col4\" class=\"data row9 col4\" >7841</td>\n",
       "      <td id=\"T_c31d0_row9_col5\" class=\"data row9 col5\" >500</td>\n",
       "      <td id=\"T_c31d0_row9_col6\" class=\"data row9 col6\" >0.97727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row10\" class=\"row_heading level0 row10\" >410</th>\n",
       "      <td id=\"T_c31d0_row10_col0\" class=\"data row10 col0\" >0.09819</td>\n",
       "      <td id=\"T_c31d0_row10_col1\" class=\"data row10 col1\" >0.81238</td>\n",
       "      <td id=\"T_c31d0_row10_col2\" class=\"data row10 col2\" >0.29742</td>\n",
       "      <td id=\"T_c31d0_row10_col3\" class=\"data row10 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row10_col4\" class=\"data row10 col4\" >8126</td>\n",
       "      <td id=\"T_c31d0_row10_col5\" class=\"data row10 col5\" >481</td>\n",
       "      <td id=\"T_c31d0_row10_col6\" class=\"data row10 col6\" >0.84889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row11\" class=\"row_heading level0 row11\" >386</th>\n",
       "      <td id=\"T_c31d0_row11_col0\" class=\"data row11 col0\" >0.09823</td>\n",
       "      <td id=\"T_c31d0_row11_col1\" class=\"data row11 col1\" >0.89506</td>\n",
       "      <td id=\"T_c31d0_row11_col2\" class=\"data row11 col2\" >0.14149</td>\n",
       "      <td id=\"T_c31d0_row11_col3\" class=\"data row11 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row11_col4\" class=\"data row11 col4\" >6380</td>\n",
       "      <td id=\"T_c31d0_row11_col5\" class=\"data row11 col5\" >494</td>\n",
       "      <td id=\"T_c31d0_row11_col6\" class=\"data row11 col6\" >0.96297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row12\" class=\"row_heading level0 row12\" >390</th>\n",
       "      <td id=\"T_c31d0_row12_col0\" class=\"data row12 col0\" >0.09838</td>\n",
       "      <td id=\"T_c31d0_row12_col1\" class=\"data row12 col1\" >0.84235</td>\n",
       "      <td id=\"T_c31d0_row12_col2\" class=\"data row12 col2\" >0.15473</td>\n",
       "      <td id=\"T_c31d0_row12_col3\" class=\"data row12 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row12_col4\" class=\"data row12 col4\" >9326</td>\n",
       "      <td id=\"T_c31d0_row12_col5\" class=\"data row12 col5\" >486</td>\n",
       "      <td id=\"T_c31d0_row12_col6\" class=\"data row12 col6\" >0.91803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row13\" class=\"row_heading level0 row13\" >409</th>\n",
       "      <td id=\"T_c31d0_row13_col0\" class=\"data row13 col0\" >0.09865</td>\n",
       "      <td id=\"T_c31d0_row13_col1\" class=\"data row13 col1\" >0.75955</td>\n",
       "      <td id=\"T_c31d0_row13_col2\" class=\"data row13 col2\" >0.35014</td>\n",
       "      <td id=\"T_c31d0_row13_col3\" class=\"data row13 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row13_col4\" class=\"data row13 col4\" >8132</td>\n",
       "      <td id=\"T_c31d0_row13_col5\" class=\"data row13 col5\" >481</td>\n",
       "      <td id=\"T_c31d0_row13_col6\" class=\"data row13 col6\" >0.88217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row14\" class=\"row_heading level0 row14\" >419</th>\n",
       "      <td id=\"T_c31d0_row14_col0\" class=\"data row14 col0\" >0.09874</td>\n",
       "      <td id=\"T_c31d0_row14_col1\" class=\"data row14 col1\" >0.79091</td>\n",
       "      <td id=\"T_c31d0_row14_col2\" class=\"data row14 col2\" >0.37138</td>\n",
       "      <td id=\"T_c31d0_row14_col3\" class=\"data row14 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row14_col4\" class=\"data row14 col4\" >2299</td>\n",
       "      <td id=\"T_c31d0_row14_col5\" class=\"data row14 col5\" >479</td>\n",
       "      <td id=\"T_c31d0_row14_col6\" class=\"data row14 col6\" >0.85834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row15\" class=\"row_heading level0 row15\" >300</th>\n",
       "      <td id=\"T_c31d0_row15_col0\" class=\"data row15 col0\" >0.09878</td>\n",
       "      <td id=\"T_c31d0_row15_col1\" class=\"data row15 col1\" >0.87206</td>\n",
       "      <td id=\"T_c31d0_row15_col2\" class=\"data row15 col2\" >0.16270</td>\n",
       "      <td id=\"T_c31d0_row15_col3\" class=\"data row15 col3\" >10</td>\n",
       "      <td id=\"T_c31d0_row15_col4\" class=\"data row15 col4\" >15714</td>\n",
       "      <td id=\"T_c31d0_row15_col5\" class=\"data row15 col5\" >490</td>\n",
       "      <td id=\"T_c31d0_row15_col6\" class=\"data row15 col6\" >0.95981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row16\" class=\"row_heading level0 row16\" >315</th>\n",
       "      <td id=\"T_c31d0_row16_col0\" class=\"data row16 col0\" >0.09901</td>\n",
       "      <td id=\"T_c31d0_row16_col1\" class=\"data row16 col1\" >0.79905</td>\n",
       "      <td id=\"T_c31d0_row16_col2\" class=\"data row16 col2\" >0.39843</td>\n",
       "      <td id=\"T_c31d0_row16_col3\" class=\"data row16 col3\" >10</td>\n",
       "      <td id=\"T_c31d0_row16_col4\" class=\"data row16 col4\" >8555</td>\n",
       "      <td id=\"T_c31d0_row16_col5\" class=\"data row16 col5\" >477</td>\n",
       "      <td id=\"T_c31d0_row16_col6\" class=\"data row16 col6\" >0.92700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row17\" class=\"row_heading level0 row17\" >389</th>\n",
       "      <td id=\"T_c31d0_row17_col0\" class=\"data row17 col0\" >0.09902</td>\n",
       "      <td id=\"T_c31d0_row17_col1\" class=\"data row17 col1\" >0.83487</td>\n",
       "      <td id=\"T_c31d0_row17_col2\" class=\"data row17 col2\" >0.17564</td>\n",
       "      <td id=\"T_c31d0_row17_col3\" class=\"data row17 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row17_col4\" class=\"data row17 col4\" >9930</td>\n",
       "      <td id=\"T_c31d0_row17_col5\" class=\"data row17 col5\" >491</td>\n",
       "      <td id=\"T_c31d0_row17_col6\" class=\"data row17 col6\" >0.94023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row18\" class=\"row_heading level0 row18\" >412</th>\n",
       "      <td id=\"T_c31d0_row18_col0\" class=\"data row18 col0\" >0.09907</td>\n",
       "      <td id=\"T_c31d0_row18_col1\" class=\"data row18 col1\" >0.82705</td>\n",
       "      <td id=\"T_c31d0_row18_col2\" class=\"data row18 col2\" >0.38552</td>\n",
       "      <td id=\"T_c31d0_row18_col3\" class=\"data row18 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row18_col4\" class=\"data row18 col4\" >7949</td>\n",
       "      <td id=\"T_c31d0_row18_col5\" class=\"data row18 col5\" >481</td>\n",
       "      <td id=\"T_c31d0_row18_col6\" class=\"data row18 col6\" >0.88007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c31d0_level0_row19\" class=\"row_heading level0 row19\" >406</th>\n",
       "      <td id=\"T_c31d0_row19_col0\" class=\"data row19 col0\" >0.09927</td>\n",
       "      <td id=\"T_c31d0_row19_col1\" class=\"data row19 col1\" >0.81869</td>\n",
       "      <td id=\"T_c31d0_row19_col2\" class=\"data row19 col2\" >0.34460</td>\n",
       "      <td id=\"T_c31d0_row19_col3\" class=\"data row19 col3\" >11</td>\n",
       "      <td id=\"T_c31d0_row19_col4\" class=\"data row19 col4\" >6951</td>\n",
       "      <td id=\"T_c31d0_row19_col5\" class=\"data row19 col5\" >484</td>\n",
       "      <td id=\"T_c31d0_row19_col6\" class=\"data row19 col6\" >0.88402</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7d469b370390>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_hyperp = study.trials_dataframe().rename(columns={'value': 'mape'})\n",
    "cols_hyperp = ['mape'] + df_hyperp.columns[df_hyperp.columns.str.contains('params_')].tolist()\n",
    "df_hyperp[cols_hyperp].sort_values('mape').head(20).style.background_gradient(cmap='coolwarm').format(precision=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e0513f3-20c9-4d43-a0aa-cc394705198f",
   "metadata": {},
   "source": [
    "Now we train a model each cutoff year (2007-2023) rather than one model every N years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f992c4d1-46a7-44fd-b38b-c55a3e351216",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_params = {'device': 'cuda',\n",
    "               'tree_method': 'hist',\n",
    "               'objective': 'reg:squarederror',\n",
    "               'colsample_bytree': 0.823,\n",
    "               'learning_rate': 0.156,\n",
    "               'max_depth': 11,\n",
    "               'min_child_weight': 3800,\n",
    "               'n_estimators': 484,\n",
    "               'subsample': 0.856\n",
    "               }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e079236d-e336-45e9-91cb-90ada885340b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5min 58s, sys: 3.9 s, total: 6min 2s\n",
      "Wall time: 5min 53s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.07626411239092291"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "cudf_reg['pred'] = cupy.nan\n",
    "\n",
    "for cutoff_year in range(first_window_final_year, current_year):\n",
    "    sel_train = cudf_reg.date.dt.year <= cutoff_year\n",
    "    sel_pred = cudf_reg.date.dt.year == cutoff_year + 1\n",
    "    reg = xgboost.XGBRegressor(**best_params)\n",
    "    reg.fit(X=cudf_reg.loc[sel_train, cols_feat_keep_low_corr], y=cudf_reg.loc[sel_train, 'mcap_daily_scaled_log'])\n",
    "    cudf_reg.loc[sel_pred, 'pred'] = reg.predict(X=cudf_reg.loc[sel_pred, cols_feat_keep_low_corr])\n",
    "\n",
    "df_score = cudf_reg.loc[~cudf_reg.pred.isnull(), ['key_daily', 'date', 'mcap_daily_scaled_log', 'pred']].to_pandas()\n",
    "mape_score(df_score.mcap_daily_scaled_log, df_score.pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "27f02d3c-8abf-4f27-8810-4d01e400ab2b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['/home/ar/data/analytics_demo/reg_xgb_best.pkl']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joblib.dump(reg, best_model_filepath)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "aa1d2e25-9fea-44d7-93af-d4d8ee1886cc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(442, 3)\n",
      "37 features contribute 90.0% of cumulative importance\n"
     ]
    },
    {
     "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>feat</th>\n",
       "      <th>imp</th>\n",
       "      <th>imp_c</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>assets_art_scaled</td>\n",
       "      <td>0.519</td>\n",
       "      <td>0.519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cashnequsd_art_scaled</td>\n",
       "      <td>0.145</td>\n",
       "      <td>0.664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ncfo_over_assetsavg_art</td>\n",
       "      <td>0.022</td>\n",
       "      <td>0.687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>intangibles_art_scaled</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>sbcomp_over_assetsavg_art</td>\n",
       "      <td>0.017</td>\n",
       "      <td>0.724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>netinccmn_over_revenue_art</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>equityavg_over_assetsavg_art</td>\n",
       "      <td>0.015</td>\n",
       "      <td>0.755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>revenueusd_art_scaled</td>\n",
       "      <td>0.015</td>\n",
       "      <td>0.770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ncfo_over_assetsavg_ary5y_avg</td>\n",
       "      <td>0.014</td>\n",
       "      <td>0.784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>cashneq_art_scaled</td>\n",
       "      <td>0.014</td>\n",
       "      <td>0.798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>opinc_over_revenue_ary5y_avg</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>debtc_over_assetsavg_art</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>ncfcommon_over_assetsavg_art</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>equityusd_art_scaled</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>ncfdiv_over_assetsavg_art</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>fcf_art_scaled</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>assets_over_assetsavg_art</td>\n",
       "      <td>0.004</td>\n",
       "      <td>0.846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>receivables_art_scaled</td>\n",
       "      <td>0.004</td>\n",
       "      <td>0.850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>prefdivis_art_scaled</td>\n",
       "      <td>0.004</td>\n",
       "      <td>0.855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>accoci_art_scaled</td>\n",
       "      <td>0.004</td>\n",
       "      <td>0.859</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             feat   imp  imp_c\n",
       "0               assets_art_scaled 0.519  0.519\n",
       "1           cashnequsd_art_scaled 0.145  0.664\n",
       "2         ncfo_over_assetsavg_art 0.022  0.687\n",
       "3          intangibles_art_scaled 0.020  0.707\n",
       "4       sbcomp_over_assetsavg_art 0.017  0.724\n",
       "5      netinccmn_over_revenue_art 0.016  0.740\n",
       "6    equityavg_over_assetsavg_art 0.015  0.755\n",
       "7           revenueusd_art_scaled 0.015  0.770\n",
       "8   ncfo_over_assetsavg_ary5y_avg 0.014  0.784\n",
       "9              cashneq_art_scaled 0.014  0.798\n",
       "10   opinc_over_revenue_ary5y_avg 0.012  0.809\n",
       "11       debtc_over_assetsavg_art 0.008  0.817\n",
       "12   ncfcommon_over_assetsavg_art 0.007  0.824\n",
       "13           equityusd_art_scaled 0.007  0.831\n",
       "14      ncfdiv_over_assetsavg_art 0.005  0.837\n",
       "15                 fcf_art_scaled 0.005  0.842\n",
       "16      assets_over_assetsavg_art 0.004  0.846\n",
       "17         receivables_art_scaled 0.004  0.850\n",
       "18           prefdivis_art_scaled 0.004  0.855\n",
       "19              accoci_art_scaled 0.004  0.859"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "reg.importance_type = 'total_gain'\n",
    "cumul_importance_thresh = 0.9\n",
    "\n",
    "df_fi = pd.DataFrame({'feat': cols_feat_keep_low_corr, 'imp': reg.feature_importances_})\n",
    "df_fi = df_fi.sort_values('imp', ascending=False).reset_index(drop=True)\n",
    "df_fi['imp_c'] = df_fi.imp.cumsum()\n",
    "important_features = df_fi.loc[df_fi.imp_c <= cumul_importance_thresh, 'feat']\n",
    "\n",
    "print(df_fi.shape)\n",
    "print(len(important_features), 'features contribute {}% of cumulative importance'.format(100.*cumul_importance_thresh))\n",
    "df_fi.to_csv(data_path.joinpath(\"analytics_demo\").joinpath('df_fi.csv'), index=None)\n",
    "display(df_fi.head(20))\n",
    "# df_fi.loc[df_fi.feat.isin(important_features), 'imp_c'].plot.line()\n",
    "df_fi.loc[:100, 'imp_c'].plot.line(figsize=(16, 8))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4f1c6e7-20a8-43f7-988d-346dd30cbcc4",
   "metadata": {},
   "source": [
    "I suppose it makes sense that asset size is the best proxy for company size. However I've never seen a single feature be assigned half the total importance. And just a half dozen features hold 75% of the predictive value!\n",
    "\n",
    "It's also surprising to see share-based compensation, as a fraction of assets (`sbcomp_over_assetsavg_art`), make an appearance in the top features. Does that perhaps indicate growth companies, which tend to have higher valuation multiples?\n",
    "\n",
    "Let's glance at prediction error."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "8f7a4720-2025-420e-aec1-f79253530351",
   "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>err</th>\n",
       "      <th>frac_err</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>820732.000</td>\n",
       "      <td>820732.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.597</td>\n",
       "      <td>0.076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.597</td>\n",
       "      <td>0.323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.195</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.433</td>\n",
       "      <td>0.039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.810</td>\n",
       "      <td>0.076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>9.565</td>\n",
       "      <td>13.799</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             err   frac_err\n",
       "count 820732.000 820732.000\n",
       "mean       0.597      0.076\n",
       "std        0.597      0.323\n",
       "min        0.000      0.000\n",
       "25%        0.195      0.017\n",
       "50%        0.433      0.039\n",
       "75%        0.810      0.076\n",
       "max        9.565     13.799"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "array([[<Axes: title={'center': 'err'}>,\n",
       "        <Axes: title={'center': 'frac_err'}>]], dtype=object)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_score['err'] = df_score.pred - df_score.mcap_daily_scaled_log\n",
    "df_score['frac_err'] = (df_score.pred - df_score.mcap_daily_scaled_log) / df_score.mcap_daily_scaled_log\n",
    "display(df_score[['err', 'frac_err']].abs().describe())\n",
    "df_score[['err', 'frac_err']].hist(bins=50, figsize=(15, 5))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9226976e-79b7-465e-b526-82f788d91954",
   "metadata": {},
   "source": [
    "Absolute percent error is more interesting to me since it's independent of company scale. We see a mean absolute percent error (known as MAPE) of `7.6%` and a median of `4%`, meaning half the predictions were within 4% of true company value. Not bad!\n",
    "\n",
    "Looks percent error (not absolute) is unsteady through time:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "f7d27fcb-55e3-4829-889d-0f787172332b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='date'>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_score[['date', 'frac_err']].groupby('date').median().sort_index().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ceca11a-23ec-4b11-b0a8-ca6c057f96ab",
   "metadata": {},
   "source": [
    "Now, notice that a positive `frac_error` means that the stock looks undervalued to XGboost. If we've done an adequate job, the higher this number goes, the better the bargain we're likely to find.\n",
    "\n",
    "So let's find out if that's true. I'll use `excess_price_chg_1y`, which measures a stock's price change minus the market's return. So a value of `-0.2` means it trailed the market by 20% over the following year, and `0.5` means it beat the market by 50% in that period. If you'd like to see how `excess_price_chg_1y` is computed, have a look at in the dbt sub-project's `models` folder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "83233424-7722-4b41-a98f-6172dd8e409b",
   "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>key_daily</th>\n",
       "      <th>date</th>\n",
       "      <th>mcap_daily_scaled_log</th>\n",
       "      <th>pred</th>\n",
       "      <th>err</th>\n",
       "      <th>frac_err</th>\n",
       "      <th>excess_price_chg_1y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LORL_2012-06-05</td>\n",
       "      <td>2012-06-05</td>\n",
       "      <td>12.757</td>\n",
       "      <td>12.285</td>\n",
       "      <td>-0.472</td>\n",
       "      <td>-0.037</td>\n",
       "      <td>0.251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SIX_2016-08-03</td>\n",
       "      <td>2016-08-03</td>\n",
       "      <td>13.701</td>\n",
       "      <td>12.666</td>\n",
       "      <td>-1.035</td>\n",
       "      <td>-0.076</td>\n",
       "      <td>0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>IDSA_2012-08-03</td>\n",
       "      <td>2012-08-03</td>\n",
       "      <td>8.516</td>\n",
       "      <td>8.434</td>\n",
       "      <td>-0.081</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>-0.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GNMK_2019-08-05</td>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>10.394</td>\n",
       "      <td>12.321</td>\n",
       "      <td>1.928</td>\n",
       "      <td>0.185</td>\n",
       "      <td>0.338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>MOBQ_2011-10-05</td>\n",
       "      <td>2011-10-05</td>\n",
       "      <td>8.746</td>\n",
       "      <td>7.371</td>\n",
       "      <td>-1.375</td>\n",
       "      <td>-0.157</td>\n",
       "      <td>-0.522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>GSIH_2009-12-03</td>\n",
       "      <td>2009-12-03</td>\n",
       "      <td>10.954</td>\n",
       "      <td>9.990</td>\n",
       "      <td>-0.964</td>\n",
       "      <td>-0.088</td>\n",
       "      <td>-0.272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PFBX_2016-04-05</td>\n",
       "      <td>2016-04-05</td>\n",
       "      <td>8.967</td>\n",
       "      <td>8.588</td>\n",
       "      <td>-0.378</td>\n",
       "      <td>-0.042</td>\n",
       "      <td>0.261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>PGI_2015-06-03</td>\n",
       "      <td>2015-06-03</td>\n",
       "      <td>11.072</td>\n",
       "      <td>11.408</td>\n",
       "      <td>0.336</td>\n",
       "      <td>0.030</td>\n",
       "      <td>0.289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>REN_2017-10-04</td>\n",
       "      <td>2017-10-04</td>\n",
       "      <td>11.324</td>\n",
       "      <td>11.225</td>\n",
       "      <td>-0.100</td>\n",
       "      <td>-0.009</td>\n",
       "      <td>0.022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>IMNPQ_2012-01-05</td>\n",
       "      <td>2012-01-05</td>\n",
       "      <td>7.336</td>\n",
       "      <td>8.537</td>\n",
       "      <td>1.201</td>\n",
       "      <td>0.164</td>\n",
       "      <td>-0.491</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          key_daily       date  mcap_daily_scaled_log   pred    err  frac_err  \\\n",
       "0   LORL_2012-06-05 2012-06-05                 12.757 12.285 -0.472    -0.037   \n",
       "1    SIX_2016-08-03 2016-08-03                 13.701 12.666 -1.035    -0.076   \n",
       "2   IDSA_2012-08-03 2012-08-03                  8.516  8.434 -0.081    -0.010   \n",
       "3   GNMK_2019-08-05 2019-08-05                 10.394 12.321  1.928     0.185   \n",
       "4   MOBQ_2011-10-05 2011-10-05                  8.746  7.371 -1.375    -0.157   \n",
       "5   GSIH_2009-12-03 2009-12-03                 10.954  9.990 -0.964    -0.088   \n",
       "6   PFBX_2016-04-05 2016-04-05                  8.967  8.588 -0.378    -0.042   \n",
       "7    PGI_2015-06-03 2015-06-03                 11.072 11.408  0.336     0.030   \n",
       "8    REN_2017-10-04 2017-10-04                 11.324 11.225 -0.100    -0.009   \n",
       "9  IMNPQ_2012-01-05 2012-01-05                  7.336  8.537  1.201     0.164   \n",
       "\n",
       "   excess_price_chg_1y  \n",
       "0                0.251  \n",
       "1                0.002  \n",
       "2               -0.360  \n",
       "3                0.338  \n",
       "4               -0.522  \n",
       "5               -0.272  \n",
       "6                0.261  \n",
       "7                0.289  \n",
       "8                0.022  \n",
       "9               -0.491  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_score = df_score.merge(df_fct_daily_targ[['key_daily', 'excess_price_chg_1y']], how='inner', on='key_daily')\n",
    "df_score.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e00dc27c-5080-47fd-a725-e1c50ddf7838",
   "metadata": {},
   "source": [
    "Let's check base performance first."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "50d0946a-b064-47eb-9e59-3242a9e67386",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "All securities, stockwise performance:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    768285.000\n",
       "mean          0.012\n",
       "median       -0.030\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "All securities, daily performance:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    184.000\n",
       "mean       0.012\n",
       "median    -0.013\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Subset to rows with 1-year return values (so they're at least a year old)\n",
    "df_score_epc = df_score[~df_score.excess_price_chg_1y.isnull()].copy()\n",
    "\n",
    "# Performance of all stocks\n",
    "print('All securities, stockwise performance:')\n",
    "display(df_score_epc.excess_price_chg_1y.agg(['count', 'mean', 'median']))\n",
    "print('\\nAll securities, daily performance:')\n",
    "display(df_score_epc[['date', 'excess_price_chg_1y']].groupby('date').mean().agg(['count', 'mean', 'median']).excess_price_chg_1y)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20314b8b-2f3c-424f-bab6-84a439a08ab4",
   "metadata": {},
   "source": [
    "Now we'll look at performance for the top couple percent of `frac_err`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "e5fd2e89-82be-4dc8-ae4e-08c3f17c7960",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 2%, stockwise performance:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    15366.000\n",
       "mean         0.370\n",
       "median      -0.197\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Top 2%, daily performance:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    184.000\n",
       "mean       0.322\n",
       "median     0.054\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Performance of top 2%\n",
    "sel_most_undervalued = df_score_epc.frac_err > df_score_epc.frac_err.quantile(.98)\n",
    "print('Top 2%, stockwise performance:')\n",
    "display(df_score_epc.loc[sel_most_undervalued, 'excess_price_chg_1y'].agg(['count', 'mean', 'median']))\n",
    "print('\\nTop 2%, daily performance:')\n",
    "display(df_score_epc.loc[sel_most_undervalued, ['date', 'excess_price_chg_1y']].groupby('date').mean().agg(['count', 'mean', 'median']).excess_price_chg_1y)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "50293f9a-7dee-4312-b45f-961603139809",
   "metadata": {},
   "source": [
    "37% stock-wise outperformance -- pretty cool, but it's cherry-picking the top 2% from 2008 to present, after the fact.\n",
    "\n",
    "Instead, what if I'd picked the 20 most-undervalued stocks each day? Since I predicted forward in time, this is actually valid methodology:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "d6e5b02c-7cd8-4431-a18a-ee0be6ba5ae5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 20 daily, stockwise performance:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    3680.000\n",
       "mean        0.612\n",
       "median     -0.112\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Top 20 daily, daily performance:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    184.000\n",
       "mean       0.612\n",
       "median     0.158\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Daily rank of prediction error, most positive is rank #1\n",
    "df_score_epc['daily_rnk'] = df_score_epc[['date', 'frac_err']].groupby('date').rank(ascending=False, method='first')\n",
    "sel_most_undervalued_daily = df_score_epc.daily_rnk <= 20\n",
    "\n",
    "print('Top 20 daily, stockwise performance:')\n",
    "display(df_score_epc.loc[sel_most_undervalued_daily, 'excess_price_chg_1y'].agg(['count', 'mean', 'median']))\n",
    "print('\\nTop 20 daily, daily performance:')\n",
    "display(df_score_epc.loc[sel_most_undervalued_daily, ['date', 'excess_price_chg_1y']].groupby('date').mean().agg(['count', 'mean', 'median']).excess_price_chg_1y)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e1008cd5-24e4-4cb0-b812-5498e5f3019c",
   "metadata": {},
   "source": [
    "Beating the market by 61.2%? Hmm... it's almost... too good to be true. I'll say more about that in the conclusion. :)\n",
    "\n",
    "What's that ride like day-to-day?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "9a0ba8a9-aa66-47bc-b57e-1a1b7d0d6096",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   184.000\n",
       "mean      0.612\n",
       "std       2.182\n",
       "min      -0.484\n",
       "25%      -0.062\n",
       "50%       0.158\n",
       "75%       0.624\n",
       "max      24.472\n",
       "Name: excess_price_chg_1y, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='date'>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_daily_perf = df_score_epc.loc[sel_most_undervalued_daily, ['date', 'excess_price_chg_1y']].groupby('date').mean()\n",
    "df_daily_perf['zero'] = 0\n",
    "display(df_daily_perf.excess_price_chg_1y.describe())\n",
    "df_daily_perf.plot.line(figsize=(15, 5))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2d5fa2a9-db6f-49c4-aa92-6ba1ae11b144",
   "metadata": {},
   "source": [
    "It's hard to tell due to the peaks, such as the GFC and the COVID crash. Let's truncate those stellar days to get a better look."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "7c96b913-5167-4674-9cf3-25d5b2568631",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='date'>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_daily_perf[df_daily_perf.excess_price_chg_1y < 3].plot.line(figsize=(15, 5))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e7db37f-29f9-4117-96bf-f888f2a5c7ef",
   "metadata": {},
   "source": [
    "Doesn't look too bad, but it also doesn't look like a fun ride for long stretches of time. Would there have been signs of upcoming underperformance? What I mean is, could this system have shown me when there weren't many deals about?\n",
    "\n",
    "We'll compute the daily median `frac_err` again, but this time only for the top 20 stocks, then we'll plot it through time. We're looking for large positive values as a sign of abundant cheap stocks, and drops as a sign that there aren't many bargains."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "9cfb00cd-568c-4327-a97e-caa18d0c0b3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='date'>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_score_epc.loc[sel_most_undervalued_daily, ['date', 'frac_err']].groupby('date').median().plot.line(figsize=(15, 5))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "597c1d07-3843-4f36-8bdd-de2436dbe7ad",
   "metadata": {},
   "source": [
    "Certainly the lower performance lately corresponds to lower median `frac_err`, but that doesn't appear universally true. Let's look at a scatter plot between the two."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "d568aa4a-8c0b-43c3-92af-7d480de4dbb6",
   "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>excess_price_chg_1y</th>\n",
       "      <th>daily_median_frac_err</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>excess_price_chg_1y</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>daily_median_frac_err</th>\n",
       "      <td>0.060</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       excess_price_chg_1y  daily_median_frac_err\n",
       "excess_price_chg_1y                  1.000                  0.060\n",
       "daily_median_frac_err                0.060                  1.000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='daily_median_frac_err', ylabel='excess_price_chg_1y'>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "df_daily_med_frac_error = df_score_epc.loc[sel_most_undervalued_daily, ['date', 'frac_err']].groupby('date').median()\n",
    "df_daily_scatter = df_daily_perf.merge(df_daily_med_frac_error, how='inner', left_index=True, right_index=True)\n",
    "df_daily_scatter = df_daily_scatter.rename(columns={'frac_err': 'daily_median_frac_err'})\n",
    "\n",
    "display(df_daily_scatter.drop(columns='zero').corr())\n",
    "sns.regplot(data=df_daily_scatter, x='daily_median_frac_err', y='excess_price_chg_1y')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c76bd44f-d4e9-46a0-bc29-bb612896deaa",
   "metadata": {},
   "source": [
    "Zooming in on the majority once again by removing the top performers:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "9b2d138c-abc8-481e-8405-37a349b65c0e",
   "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>excess_price_chg_1y</th>\n",
       "      <th>daily_median_frac_err</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>excess_price_chg_1y</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>daily_median_frac_err</th>\n",
       "      <td>0.047</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       excess_price_chg_1y  daily_median_frac_err\n",
       "excess_price_chg_1y                  1.000                  0.047\n",
       "daily_median_frac_err                0.047                  1.000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='daily_median_frac_err', ylabel='excess_price_chg_1y'>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sel_remove_outliers = df_daily_scatter.excess_price_chg_1y < 4\n",
    "display(df_daily_scatter[sel_remove_outliers].drop(columns='zero').corr())\n",
    "sns.regplot(data=df_daily_scatter[sel_remove_outliers], x='daily_median_frac_err', y='excess_price_chg_1y')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2fb03d82-1033-4e41-9f86-a907be04f95f",
   "metadata": {},
   "source": [
    "Yeah, no strong relationship evident there. In other words, it probably wouldn't make sense to wait for days with tons of cheap-looking stocks. Anyway, probably only happens during a big crash, precisely when an instrument like this isn't needed to infer that stocks are cheap."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b3a59c8a-a103-40fd-a294-794df654d8ff",
   "metadata": {},
   "source": [
    "# Conclusion\n",
    "\n",
    "That's a lot of progress in one notebook! However, the preceding is far, far from anything you'd want to use to inform investing decisions. Here's a list of unsolved sub-problems in no particular order.\n",
    "\n",
    "- Dataset\n",
    "    - Testing\n",
    "        - missing\n",
    "        - conformity to expectations\n",
    "    - Filter unwanted instances\n",
    "        - International stocks\n",
    "        - Meme stocks/penny stocks\n",
    "    - Ensure liquidity\n",
    "        - Is it actually traded?\n",
    "        - Market cap floor\n",
    "- Modeling / problem framing\n",
    "    - Need better features\n",
    "        - Typically, one should study which features are working and try to \"repeat what works\", as Charlie Munger said.\n",
    "        - Cohort/industry/competition level metrics\n",
    "        - Lagged target\n",
    "        - Do a better job capturing trends in financials\n",
    "        - Capture insider trading & other corporate events\n",
    "        - Measure \"classic factors\" directly: dividends, quality, momentum, value\n",
    "    - Try other models\n",
    "    - Try model averaging and/or stacking\n",
    "    - Consider targeteting market outperformance directly, rather than mcap\n",
    "    - Handle market system changes\n",
    "        - e.g., credit cycle: firm debt sensitivity is higher now than in much of this dataset\n",
    "    - Unsupervised ML as preprocessing: See if there are clusters which, if you model each separately, improve predictions.\n",
    "    - Should you quantify uncertainty? That is, use statistics more clasically?\n",
    "- Understanding performance\n",
    "    - Weighting by estimated (under-) valuation rather than equal weighting.\n",
    "    - Optimal purchase strategy -- buy every day or wait for good deals?\n",
    "    - Optimal number of stocks to buy\n",
    "    - Multi-period scoring: how long to hold?\n",
    "    - Exit strategy\n",
    "    - Taxation considerations\n",
    "    - Compounding performance -- Is there a risk of ruin?\n",
    "    - Industry/sector diversification\n",
    "    - Systematic error analysis: Find and understand any *error trends*; try to engineer features or acquire data to address them."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}