notebooks/4-creating-custom-xg-pipelines.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Custom xG Models\n",
    "This notebook trains two custom xG pipelines for shots from open play using Logistic Regression and Gradient Boosting Probability Trees on the following set of features:\n",
    "\n",
    "`start_x_a0` :\n",
    "    The x-coordinate of the shot  \n",
    "`start_y_a0` :\n",
    "    The y-coordinate of the shot  \n",
    "`shot_angle_a0` :\n",
    "    The angle of the shot  \n",
    "`start_dist_to_goal_a0` :\n",
    "    The distance to the center of the goal  \n",
    "`bodypart_id_a0` :\n",
    "    Bodypart used to take the shot (foot or head/other body part)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import itertools\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "from soccer_xg.api import DataApi\n",
    "import soccer_xg.xg as xg\n",
    "from soccer_xg.utils import sample_temporal"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Config\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# dataset\n",
    "dir_data = \"../data\"\n",
    "provider = 'wyscout_opensource'\n",
    "leagues = ['ENG', 'ESP', 'ITA', 'GER', 'FRA']\n",
    "seasons = ['1718']\n",
    "\n",
    "# features\n",
    "store_features = f'../data/{provider}/features.h5'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Optionally, you can use dask to speed up the parameter tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border: 2px solid white;\">\n",
       "<tr>\n",
       "<td style=\"vertical-align: top; border: 0px solid white\">\n",
       "<h3 style=\"text-align: left;\">Client</h3>\n",
       "<ul style=\"text-align: left; list-style: none; margin: 0; padding: 0;\">\n",
       "  <li><b>Scheduler: </b>inproc://134.58.41.109/18671/1</li>\n",
       "  <li><b>Dashboard: </b><a href='http://134.58.41.109:8787/status' target='_blank'>http://134.58.41.109:8787/status</a></li>\n",
       "</ul>\n",
       "</td>\n",
       "<td style=\"vertical-align: top; border: 0px solid white\">\n",
       "<h3 style=\"text-align: left;\">Cluster</h3>\n",
       "<ul style=\"text-align: left; list-style:none; margin: 0; padding: 0;\">\n",
       "  <li><b>Workers: </b>1</li>\n",
       "  <li><b>Cores: </b>4</li>\n",
       "  <li><b>Memory: </b>2.00 GB</li>\n",
       "</ul>\n",
       "</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<Client: 'inproc://134.58.41.109/18671/1' processes=1 threads=4, memory=2.00 GB>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import dask\n",
    "dask.config.set({'temporary_directory': '/tmp'})\n",
    "\n",
    "cluster = False\n",
    "#cluster = '134.58.41.90:8786'\n",
    "\n",
    "if not cluster:\n",
    "    from dask.distributed import Client\n",
    "    client = Client(processes=False, threads_per_worker=4,\n",
    "                    n_workers=1, memory_limit='2GB')\n",
    "    \n",
    "else:\n",
    "    from dask.distributed import Client\n",
    "    client = Client(cluster)\n",
    "    \n",
    "client"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing\n",
    "We start by loading the raw SPADL data and precomputed features. Make sure to run the `1-load-and-convert-wyscout-data` and `3-computing-and-storing-features` notebooks first to download the data set, convert it to the SPADL format and compute the xG features.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "api = DataApi([f\"{dir_data}/{provider}/spadl-{provider}-{l}-{s}.h5\" \n",
    "        for (l,s) in itertools.product(leagues, seasons)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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></th>\n",
       "      <th>type_id_a0</th>\n",
       "      <th>type_id_a1</th>\n",
       "      <th>type_id_a2</th>\n",
       "      <th>bodypart_id_a0</th>\n",
       "      <th>bodypart_id_a1</th>\n",
       "      <th>bodypart_id_a2</th>\n",
       "      <th>result_id_a1</th>\n",
       "      <th>result_id_a2</th>\n",
       "      <th>start_x_a0</th>\n",
       "      <th>start_y_a0</th>\n",
       "      <th>...</th>\n",
       "      <th>speed_a02</th>\n",
       "      <th>shot_angle_a0</th>\n",
       "      <th>shot_angle_a1</th>\n",
       "      <th>shot_angle_a2</th>\n",
       "      <th>caley_zone_a0</th>\n",
       "      <th>caley_zone_a1</th>\n",
       "      <th>caley_zone_a2</th>\n",
       "      <th>angle_zone_a0</th>\n",
       "      <th>angle_zone_a1</th>\n",
       "      <th>angle_zone_a2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>game_id</th>\n",
       "      <th>action_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">2500098</th>\n",
       "      <th>17</th>\n",
       "      <td>shot</td>\n",
       "      <td>dribble</td>\n",
       "      <td>cross</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>success</td>\n",
       "      <td>success</td>\n",
       "      <td>99.75</td>\n",
       "      <td>26.52</td>\n",
       "      <td>...</td>\n",
       "      <td>1.319750</td>\n",
       "      <td>0.499778</td>\n",
       "      <td>0.483780</td>\n",
       "      <td>0.065500</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>12</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>shot</td>\n",
       "      <td>corner_crossed</td>\n",
       "      <td>pass</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>success</td>\n",
       "      <td>fail</td>\n",
       "      <td>91.35</td>\n",
       "      <td>35.36</td>\n",
       "      <td>...</td>\n",
       "      <td>4.249364</td>\n",
       "      <td>0.517985</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.363334</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>12</td>\n",
       "      <td>21</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>shot</td>\n",
       "      <td>clearance</td>\n",
       "      <td>cross</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>fail</td>\n",
       "      <td>fail</td>\n",
       "      <td>75.60</td>\n",
       "      <td>29.92</td>\n",
       "      <td>...</td>\n",
       "      <td>5.874066</td>\n",
       "      <td>0.242481</td>\n",
       "      <td>0.491555</td>\n",
       "      <td>0.043863</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>12</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>shot</td>\n",
       "      <td>cross</td>\n",
       "      <td>dribble</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>success</td>\n",
       "      <td>success</td>\n",
       "      <td>92.40</td>\n",
       "      <td>43.52</td>\n",
       "      <td>...</td>\n",
       "      <td>2.061543</td>\n",
       "      <td>0.371538</td>\n",
       "      <td>0.134860</td>\n",
       "      <td>0.167545</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>15</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>shot</td>\n",
       "      <td>pass</td>\n",
       "      <td>pass</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>foot</td>\n",
       "      <td>success</td>\n",
       "      <td>success</td>\n",
       "      <td>99.75</td>\n",
       "      <td>37.40</td>\n",
       "      <td>...</td>\n",
       "      <td>1.289870</td>\n",
       "      <td>0.978291</td>\n",
       "      <td>0.654611</td>\n",
       "      <td>0.320841</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 59 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  type_id_a0      type_id_a1 type_id_a2 bodypart_id_a0  \\\n",
       "game_id action_id                                                        \n",
       "2500098 17              shot         dribble      cross           foot   \n",
       "        40              shot  corner_crossed       pass           foot   \n",
       "        77              shot       clearance      cross           foot   \n",
       "        140             shot           cross    dribble           foot   \n",
       "        145             shot            pass       pass           foot   \n",
       "\n",
       "                  bodypart_id_a1 bodypart_id_a2 result_id_a1 result_id_a2  \\\n",
       "game_id action_id                                                           \n",
       "2500098 17                  foot           foot      success      success   \n",
       "        40                  foot           foot      success         fail   \n",
       "        77                  foot           foot         fail         fail   \n",
       "        140                 foot           foot      success      success   \n",
       "        145                 foot           foot      success      success   \n",
       "\n",
       "                   start_x_a0  start_y_a0  ...  speed_a02  shot_angle_a0  \\\n",
       "game_id action_id                          ...                             \n",
       "2500098 17              99.75       26.52  ...   1.319750       0.499778   \n",
       "        40              91.35       35.36  ...   4.249364       0.517985   \n",
       "        77              75.60       29.92  ...   5.874066       0.242481   \n",
       "        140             92.40       43.52  ...   2.061543       0.371538   \n",
       "        145             99.75       37.40  ...   1.289870       0.978291   \n",
       "\n",
       "                   shot_angle_a1  shot_angle_a2  caley_zone_a0  caley_zone_a1  \\\n",
       "game_id action_id                                                               \n",
       "2500098 17              0.483780       0.065500              2              3   \n",
       "        40              0.000000       0.363334              3              8   \n",
       "        77              0.491555       0.043863              6              3   \n",
       "        140             0.134860       0.167545              4              5   \n",
       "        145             0.654611       0.320841              1              3   \n",
       "\n",
       "                   caley_zone_a2  angle_zone_a0  angle_zone_a1  angle_zone_a2  \n",
       "game_id action_id                                                              \n",
       "2500098 17                     8              9             12             18  \n",
       "        40                     4             12             21             12  \n",
       "        77                     0             18             12             18  \n",
       "        140                    0             12             15             18  \n",
       "        145                    4              6              9             15  \n",
       "\n",
       "[5 rows x 59 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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></th>\n",
       "      <th>goal</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>game_id</th>\n",
       "      <th>action_id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">2500098</th>\n",
       "      <th>17</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    goal\n",
       "game_id action_id       \n",
       "2500098 17         False\n",
       "        40         False\n",
       "        77         False\n",
       "        140        False\n",
       "        145        False"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "features = []\n",
    "labels = []\n",
    "# load features and labels for each season\n",
    "for (l,s) in itertools.product(leagues, seasons):\n",
    "    features.append(pd.read_hdf(store_features, key=f'{l}/{s}/features'))\n",
    "    labels.append(pd.read_hdf(store_features, key=f'{l}/{s}/labels'))\n",
    "features = pd.concat(features)\n",
    "labels = pd.concat(labels)\n",
    "# filter open play shots\n",
    "open_play_idx = features[features.type_id_a0 == 'shot'].index\n",
    "features = features.loc[open_play_idx]\n",
    "labels = labels.loc[open_play_idx]\n",
    "\n",
    "display(features.head())\n",
    "display(labels.to_frame().head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split train - test\n",
    "Create a train, validation (used for calibration) and test set (used for evalution)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "games_train, games_val, games_test = sample_temporal(api, size_val=0.2, size_test=0.3)\n",
    "\n",
    "X_train, y_train = features.loc[games_train], labels.loc[games_train]\n",
    "data_train = X_train.assign(goal=y_train)\n",
    "X_val, y_val = features.loc[games_val], labels.loc[games_val]\n",
    "data_val = X_train.assign(goal=y_val)\n",
    "X_test, y_test = features.loc[games_test], labels.loc[games_test]\n",
    "data_test = X_test.assign(goal=y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dataset = train\n",
      "shape = (20652, 59)\n",
      "--------------------\n",
      "dataset = val\n",
      "shape = (7781, 59)\n",
      "--------------------\n",
      "dataset = test\n",
      "shape = (11965, 59)\n",
      "--------------------\n"
     ]
    }
   ],
   "source": [
    "for name, X in [(\"train\", X_train), (\"val\", X_val), (\"test\", X_test)]:\n",
    "    print(\"dataset =\", name)\n",
    "    print(\"shape =\", X.shape)\n",
    "    print(\"-\" * 20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Learning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature selection\n",
    "Note that `a0` corresponds to the shot itself, `a1` would be the action before the shot, `a2` the action before that, etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "NUM = ['start_x_a0', 'start_y_a0', 'start_dist_to_goal_a0', 'shot_angle_a0']  # numerical features\n",
    "CAT = ['bodypart_id_a0']  # categorical features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "%autoreload 2\n",
    "from soccer_xg.ml import logreg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_logreg = logreg.logreg_gridsearch_classifier(\n",
    "    numeric_features=NUM,\n",
    "    categoric_features=CAT,\n",
    "    use_dask=True,\n",
    "    scoring='neg_brier_score',\n",
    "    learning_rate='optimal',\n",
    "    n_iter=1000\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/pieterr/Jupiter/Projects/soccer_dataprovider_comparison/.venv/lib/python3.6/site-packages/sklearn/model_selection/_search.py:281: UserWarning: The total space of parameters 400 is smaller than n_iter=1000. Running 400 iterations. For exhaustive searches, use GridSearchCV.\n",
      "  % (grid_size, self.n_iter, grid_size), UserWarning)\n"
     ]
    }
   ],
   "source": [
    "search_logreg.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.07954030983164834"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "search_logreg.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'max_dev': 43.84497552825291, 'residual_area': 766.1075845323285, 'roc': 0.781085666916153, 'brier': 0.08463387301843504, 'ece': 0.012483117426868794, 'ace': 0.012690942667727058, 'fig': <Figure size 720x360 with 3 Axes>}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = xg.OpenplayXGModel()\n",
    "model.model = search_logreg.best_estimator_\n",
    "model._training_seasons = [\"1718\"]\n",
    "model._fitted = True\n",
    "print(model.validate(data_test, validation_seasons=[\"1718\"]))\n",
    "model.save_model(f'{provider}_openplay_logreg_basic')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bodypart_id_a0</th>\n",
       "      <td>-3.017892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bodypart_id_a0</th>\n",
       "      <td>-2.345470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>start_dist_to_goal_a0</th>\n",
       "      <td>-0.683558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>start_y_a0</th>\n",
       "      <td>-0.012526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>start_x_a0</th>\n",
       "      <td>0.044673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>shot_angle_a0</th>\n",
       "      <td>0.487839</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       importance\n",
       "bodypart_id_a0          -3.017892\n",
       "bodypart_id_a0          -2.345470\n",
       "start_dist_to_goal_a0   -0.683558\n",
       "start_y_a0              -0.012526\n",
       "start_x_a0               0.044673\n",
       "shot_angle_a0            0.487839"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = xg.XGModel.load_model(f'{provider}_logreg_basic')\n",
    "(pd.DataFrame(\n",
    "    data={'importance': model.model.named_steps['clf'].coef_[0]}, \n",
    "    index=CAT*2 + NUM)\n",
    "    .sort_values(by='importance'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "%autoreload 2\n",
    "from soccer_xg.ml import xgboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_xgb = xgboost.xgboost_gridsearch_classifier(\n",
    "    numeric_features=NUM,\n",
    "    categoric_features=CAT,\n",
    "    use_dask=True,\n",
    "    scoring='neg_brier_score',\n",
    "    n_iter=1000\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_xgb.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.0796168605722396"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "search_xgb.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'max_dev': 30.960270602527984, 'residual_area': 857.1900614979984, 'roc': 0.7823261149160066, 'brier': 0.08429739509683612, 'ece': 0.005332297751757049, 'ace': 0.008145586900316519, 'fig': <Figure size 720x360 with 3 Axes>}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = xg.OpenplayXGModel()\n",
    "model.model = search_xgb.best_estimator_\n",
    "model._training_seasons = [\"1718\"]\n",
    "model._fitted = True\n",
    "print(model.validate(data_test, validation_seasons=[\"1718\"]))\n",
    "model.save_model(f'{provider}_openplay_xgboost_basic')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Post calibration\n",
    "Let's see if we can improve the model with post-calibration..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "%autoreload 2\n",
    "from soccer_xg.calibration import CalibratedClassifierCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>max_dev</th>\n",
       "      <th>residual_area</th>\n",
       "      <th>roc</th>\n",
       "      <th>brier</th>\n",
       "      <th>ece</th>\n",
       "      <th>ace</th>\n",
       "      <th>fig</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>method</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>isotonic</th>\n",
       "      <td>39.564576</td>\n",
       "      <td>709.146478</td>\n",
       "      <td>0.781733</td>\n",
       "      <td>0.084732</td>\n",
       "      <td>0.008976</td>\n",
       "      <td>0.005764</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta</th>\n",
       "      <td>39.564332</td>\n",
       "      <td>709.137344</td>\n",
       "      <td>0.781733</td>\n",
       "      <td>0.084732</td>\n",
       "      <td>0.008976</td>\n",
       "      <td>0.005764</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta_ab</th>\n",
       "      <td>39.564904</td>\n",
       "      <td>709.146603</td>\n",
       "      <td>0.781733</td>\n",
       "      <td>0.084732</td>\n",
       "      <td>0.008976</td>\n",
       "      <td>0.005764</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta_am</th>\n",
       "      <td>39.564701</td>\n",
       "      <td>709.146401</td>\n",
       "      <td>0.781733</td>\n",
       "      <td>0.084732</td>\n",
       "      <td>0.008976</td>\n",
       "      <td>0.006127</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>base</th>\n",
       "      <td>30.960271</td>\n",
       "      <td>857.190061</td>\n",
       "      <td>0.782326</td>\n",
       "      <td>0.084297</td>\n",
       "      <td>0.005332</td>\n",
       "      <td>0.008269</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sigmoid</th>\n",
       "      <td>45.334279</td>\n",
       "      <td>1358.774246</td>\n",
       "      <td>0.782326</td>\n",
       "      <td>0.085594</td>\n",
       "      <td>0.022263</td>\n",
       "      <td>0.026060</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            max_dev  residual_area       roc     brier       ece       ace  \\\n",
       "method                                                                       \n",
       "isotonic  39.564576     709.146478  0.781733  0.084732  0.008976  0.005764   \n",
       "beta      39.564332     709.137344  0.781733  0.084732  0.008976  0.005764   \n",
       "beta_ab   39.564904     709.146603  0.781733  0.084732  0.008976  0.005764   \n",
       "beta_am   39.564701     709.146401  0.781733  0.084732  0.008976  0.006127   \n",
       "base      30.960271     857.190061  0.782326  0.084297  0.005332  0.008269   \n",
       "sigmoid   45.334279    1358.774246  0.782326  0.085594  0.022263  0.026060   \n",
       "\n",
       "           fig  \n",
       "method          \n",
       "isotonic  None  \n",
       "beta      None  \n",
       "beta_ab   None  \n",
       "beta_am   None  \n",
       "base      None  \n",
       "sigmoid   None  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = []\n",
    "\n",
    "for method in [None, 'sigmoid', 'isotonic', 'beta', 'beta_am', 'beta_ab']:\n",
    "    # calibrate\n",
    "    calibrated_model = CalibratedClassifierCV(model.model, method=method, cv='prefit')\n",
    "    calibrated_model.fit(X_val, y_val);\n",
    "    # store\n",
    "    model = xg.OpenplayXGModel()\n",
    "    model.model = calibrated_model\n",
    "    model._training_seasons = [\"1718\"]\n",
    "    model._fitted = True\n",
    "    # validate\n",
    "    method = 'base' if method is None else method\n",
    "    validation = model.validate(data_test, validation_seasons=[\"1718\"], plot=False)\n",
    "    validation['method'] = method\n",
    "    results.append(validation)\n",
    "\n",
    "pd.DataFrame(results).set_index('method').sort_values(by='ace')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Apparently, post-calibration does not help here."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## xG heatmaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "%autoreload 2\n",
    "import soccer_xg.features as fn\n",
    "from soccer_xg.visualisation import plot_heatmap"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shots by foot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "start = np.array(list(itertools.product(np.arange(0, 106, 1), np.arange(0, 69, 1))))\n",
    "df = pd.DataFrame({'start_x': start[:,0], 'start_y': start[:,1], 'bodypart_id': 'foot'})\n",
    "df = pd.concat([\n",
    "    fn.startlocation(df), \n",
    "    fn.startpolar(df),\n",
    "    fn.bodypart(df),\n",
    "    fn.goalangle(df),\n",
    "], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQUAAAI+CAYAAAC1wx9iAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3de5xkZX3n8e9T1bfp6RlmhukZBgZEQQzecYXsJl6INxSJEXGNuMGAq0LUGI1R4+pGiURf6npZ1yisUaIoiAuKNxQvcWLMDYiDbrKbKAgyMjPMnZm+d1U9+eOcnqlfz3me7qo5fepU9ef9es1ruuv0udSpqm+den71PI/z3gsA5lQ6fQAAyoVQAGAQCgAMQgGAQSgAMAgFAMayCAXn3NXOuf/exnqnOOfGnHPVpTiuMnHOfdM597udPo6FOOcudc79sNPH0ctKFwrOufucc8/Kc5ve+yu89+9udd/e+/u99yPe+3qex1NG3vvnee8/U/R+nXOPdM59wTm32zl30Dn3M+fc/3LObe7AseT+3OtGpQuFXuSc68t5ez1x5eKcO13SP0raLuks7/1qSb8u6R5JT+nksS1r3vtS/ZN0n6RnZdw+KOkjSp5A29OfB5uWv0XSjnTZKyV5Saeny/5S0lXpz+slfV3SAUn7JP2NknC8TlJD0qSksXR7p6bb6UvXXSfp2nQf+yXdErgPl0r6W0kflrRX0lXp8f8PSfdLelDS1ZJWtHD8n5B0q6RxSc+SdKKkmyXtlnSvpNc3bescSXdKOpju60Pp7UOSPpce0wFJd0jamC7bIumV6c8VSe+Q9AtJuyR9VtJx6bK5c/K76X3ZI+ntgfMwIOkuSb+f/l5Nz8ufpL9/TtLXWnx+XCrph+m53J/e9+c1LT9O0qfSc/lAeu6r6bLTJP1Vev/3SPq8pDXpstjjf5mkben+rpB0tqSfpOfwY037Dm6/6bn9Nkn/L93WtZKGOv2aO+ocd/oAWgiFP5X0D5I2SBqV9HeS3p0ue66knZIeI2k4fbKFQuG9Sl6Q/em/p0pyWfvW0aHwDUk3Slqbrvv0yBO3Jun3JfVJWqEkIL6qJFhWSfqapPe2cPwPKXkXraR/80+S/kTJC+8Rkn4u6bz07/9e0iXpzyOS/mP68+XpfoeVvED/g6TVGaHwCkl3p9sdkfQlSdfNOyefTO/XEyRNSzozcC4em74AzpT09vQxnHuR7pR0aRuhMCvpVel9+D0lQTr3GH5Z0jWSVqbPldslXZ4uO13Ss5UE9KikH0j6SOi513Rfr1YSqM+RNCXplnTbJykJzae3sP1/lnRy+jz4W6XPyzL96/gBtBAK90g6v+n38yTdl/78aaUvsKYHJxQKfyrpK3PLYvtuelL0Sdqk5J1k7SKfuPc3/e6UvMOf1nTbf5J0bwvH/9mm5b/avP30trdJujb9+QeSrpS0ft7fvEJJmD4+45i36EgofE/Sa5qWPUrJC7Gv6Zxsblp+u6SXRs7HmyT9m5JweGTT7TVJz236/XVK3n3HJH0ycm7vbvp9OD2eEyRtVBJQzVdgF0v6fmBbL5S0dRGP/0lNt+2V9NtNv98s6Q0tbP+Kpt/Pl3RPp19z8/91U5vCiUouZ+f8Ir1tbtm2pmXNP8/3ASXvgt92zv3cOffHi9z/yZL2ee/3L/Lvm49hVOm7u3PugHPugKRvpbdLizv+5tseJunEuW2l2/tvSl4UkvRfJZ0h6V+dc3c45y5Ib79O0m2SvuCc2+6ce79zrj9jX1nnuq9p+1LyLj9nQskVRchn0mO+1Xv/s6bb9yoJW0mS9/5j3vs1Sj4aZh3XUfv23k+kP46k++iXtKPpvFyj5F1dzrmNaaPmA865g0quyNZH9jPnwaafJzN+H2lh+82PY/NzuDS6KRS2K3nQ55yS3iYlnx+bW6tPDm3Ee3/Ie/8m7/0jJL1A0h865545tziy/22S1jnn1izyeJu3tUfJk+cx3vs16b/jvPdzL6TFHH/z9rYpucpY0/Rvlff+/PQ+/sx7f7GSF8P7JN3knFvpvZ/13l/pvX+0pF+TdIGkl2fsK+tc12RfDK34uJJ2nPOcc80NiN+T9KI2t5llm5IrhfVN52W19/4x6fL3KDmPj/NJo+bvKLmKm3OsXYYX2r5kH9vm53BplDUU+p1zQ03/+iTdIOkdzrlR59x6JZ+nP5f+/RclXeacO9M5Nywp+J0E59wFzrnTnXNOyef0upKPBVLypH9E1nre+x2Svinp4865tc65fufc0xZzZ7z3DSWfwT/snJt71zrJOXdeq8eful3SIefcW51zK5xzVefcY51zZ6fb/h3n3Gi63wPpOg3n3G845x6XVi8OKvlI0MjY/g2S3uice7hzbkTJk/1G731tMfe3mXPuEiVtF5dKer2kz6TblKR3SXqqc+5DzrmT0r9fr6T9oWXpY/RtSR90zq12zlWcc6c5556e/skqJR9NHkr39+Z5mwg+/ou00PYl6bXOuc3OuXVK2lhuPIb9LYmyhsKtSt5Z5/69S0kr8p1KWn3/r6QfpbfJe/9NSR+V9H0lHw3+Id3OdMa2Hynpu0oevL+X9HHv/ffTZe9VEjwHnHN/lLHuJUpeSP+qpIHpDS3cp7fOHVt6afldJZ/VWz1++eR7ExdIeqKS1vc9kv5CScu7lDRc/otzbkzS/1TyeX9Syefum5QEwv+X9NdKPlLM9+n09h+k259S0mjaEufcKUo+Crzcez/mvb9eyWP44fR+/FRJ+8hmST92zh1S0vi2XQsHY8jLlTS+zrXw36QjH1GulPQkJW8G31DSgNpsocd/IQttX5KuVxJcP1fSTnZVG/tZUnMttj3FOXemklbewXbe3Tqt248f2Zxz9ylpzP1up48lpqxXCi1zzl3onBt0zq1V8jn6a930gur240fv6JlQUFKD36XkkqyupH7dTbr9+NEjevLjA4D29dKVAoAcEAoADEIBgEEoADAIBQAGoQDAIBQAGIQCAINQAGAQCgAMQgGAQSgAMAgFAAahAMAgFAAYhAIAg1AAYBAKAAxCAYBBKAAwCAUABqEAwCAUABiEAgCDUABgEAoADEIBgNG3wHImmgR6kwst4EoBgEEoADAIBQAGoQDAIBQAGIQCAINQAGAQCgAMQgGAQSgAMAgFAAahAMAgFAAYhAIAg1AAYBAKAAxCAYBBKAAwCAUABqEAwCAUABiEAgCDUABgEAoAjOhkMOf+2ooFN8BsMfnhXCLknntnteGEx2nr1q1Lvi+uFIAuMD7R0O7duwvZV/RK4a++tHnBDdR9I7eDWe4aXCsg4NkX7VBl4PRC9sWVAgCDUABgEAoAjIWmopdEu0G7aCPIT0PL/TlY3HOJKwUABqEAwCAUABiEAgCDUABgEAoAjGhJklLkwig7Ui7sNVwpADAIBQAGoQDAIBQAGIQCAGNRHaKWi+VSSaBagBiuFAAYhAIAg1AAYBAKAAxCAYBBKAAwlmVJspdKj5QXpbrvncdTkqrOdXT/XCkAMAgFAAahAMAgFAAYhAIAo2erD91YYei1SkKvVQWKknXeijyTXCkAMAgFAAahAMAgFAAYhAIAg1AAYHR9SbLMpcduLDFSRgRXCgAMQgGAQSgAMAgFAAahAMAoVfWhzJWEkDJXGMpcSSjzeSun4h5LrhQAGIQCAINQAGAQCgAMQgGAUarqQ5l1urW805WETt9/FIcrBQAGoQDAIBQAGIQCAINQAGAQCgCMwkuSZe70VFTZjfKiVC/x86CMmCEKQMcQCgAMQgGAQSgAMAgFAMaSVR86XWUoRQt7AVWGwiomJa4WNEo87FzFuU4fQsu4UgBgEAoADEIBgEEoADAIBQAGoQDAOOaS5HIvPRbVuSnP+1lUebHMpcKi5HUOijyVXCkAMAgFAAahAMAgFAAYhAIAY1HVh05XGMogzyoDlYRyd7Ba7rhSAGAQCgAMQgGAQSgAMAgFAEa0+lCWqkO3TtJS1ipDmSsJnR9Erxhlfjcu87EB6ABCAYBBKAAwCAUABqEAwCAUABhLNkNUO7qx9NjpsmO75cVzz9ulvXsbevipx/4UKKJw7UtSHs+LU2szR935TzMaGLhziY7G4kphmdq7t6Hx8d56ofWyRkOanZ0tZF+lulJAceauEL5603pze1tXK7kcUVxRA+QWpdriHJOPOnOnXGXFEh2NxZUCAINQAGAQCgCMwtsUurHCIHVnlWEx+5n/N3k+Onk+BvXctlQSLZ4bL7VYr2gfVwoADEIBgEEoADAIBQAGoQDAIBQAGF3/Neeydm6SWi89LlXZ8aj96EgnpsXe43bOcztlxOUyRmOZcaUAwCAUABiEAgCDUABgEAoAjCWrPuTagYjOTcH9tHNP6t4fHt5ssec2Vklo7xjaWCnH/bcj13fQono3tYErBQAGoQDAIBQAGIQCAINQAGAQCgCMYy5JFjXmYqcVUXqM7aPd0mP2frJ/jontv53yYqv3p16CGl47HbyqoVJy5Jx1+p260/sHUDKEAgCDUABgEAoADEIBgFGq4dg6PbRat3ZuCu8nW3vDsYWXhbbRTsWg7rPXaZSg+hBSiTxvQueg6mLPm87OsM2VAgCDUABgEAoADEIBgEEoADAWVX0oc/+G5V5laGeYtLo/0r49v6pQRCUh2U/2snb204jspwiVSCUh1Pch+hTscKGFKwUABqEAwCAUABiEAgCDUABgEAoAjMI7ROU921OntVPeDG6rzc5N4e0tvM789UMlwbzLi6EyYnD/bZUq833Pq7jss12NPG6NwPMjVsbMKn8X+arhSgGAQSgAMAgFAAahAMAgFAAY0epDWTpClaHTU1vbC20r585NLa8jJ5+25s9v1Q9VGWYj7x+tVhJiy0IVg/i2stfJvfoQOKPVyPMz1CGqP/ZoB6ocReFKAYBBKAAwCAUABqEAwCAUABiEAgBjyTpElbnjUzvjLQa3FVnWaunxWDs3Hb2fcOcm3/Rzs1DpMTYO4kwbJcEZVVtaJ1qSDKzTyPk9r+IDJclICbHfBR7t8r48uFIAYBEKAAxCAYBBKAAwCAUARuHDscUU1fGpiG0VJd4qH+7c5Jt+bhaqMoQqDJI067MrCbM+/PQKHXdonZnAPqRwlSFUlYjtPybYucnVgusMBKoPDTfb8v6LwpUCAINQAGAQCgAMQgGAQSgAMEpVfchTnv0bpNaHVpNa7+PQ1tBqbUzS0vBOStebX20IVRlCFYZkWesVg9lA34epxkDL+w/tJ9b3IVaZCAn1cQj2b5A0FKoyRHafNekMk8EA6JievVIAjsW7L9yqQ/tmteGUocO3ucBFmYu8j1cCy2I9K7MGiB0f86q4Q8F18kQoABkO7ZvV9ERsfO2CeckX9CGCUAAyzF0h/NF1jz98W55tCisr08F1hjOWPfP0u+UK+rRPmwIAg1AAYBzzx4cyD7vWbdoZWi1UdpSk2cjQZr7pZ7O9wDqxzk2hkuCU7w+uE1oW2s9Uo51txcqo8eHg5o5jT23V4WVZpUJJGop0iMr6KLCQWAerInClAMAgFAAYhAIAg1AAYBAKAIzCv7zU6SHXYtvLc2KX2PZiHZ/C+w9M7BIbji045FlVPl02vxU+NElLrHNTqPU/Wn0IVBOmfHaHqIlARylJmqgPZt4+HamYhPZfayTvk3P3d9fMqsy/a9ZXCT9z1vWNL7j+fLEvNhWBbzQuQ9577X6wJqrJyMLHh2VmaqqhP3vzHu15sK7Va3j4cTSeFcvIgf11Xf6SXarNep3x2AENDfHw42g8K5aJPbvquuK3d+mcXx/SOz8yqmql9SHOsTwQCsvAzgdquvwlD+q83xrWa9+6Ri40MAAgQqHn3X/vrC5/yYN68SUjuuy1x3X6cNAFqD50QLBUGSkvRsdbDPjZT2f1B5fs0qveeJxe8NJVh8unsQ5RwdmeAqXKZJ3WOzFNNLLLiKHbD9WHMm+XpLFASXI8cLskjdWyS5xT9eSYpxvJfdo2ufbwsrly5XyxkuS+/pWZt9dXhN+PR/sOBpcVgVDoUQ/uqOkNL9+l171tjc574UinDwddhI8PPWhyoqE3vnKPXnLZKgIBLSMUekyj4fXON+3T6Y/q13959epOHw66EB8fesw1HzmoPbvquvr6DeHhh4EIQqGH3PbVCX3j5nF99isbNTDoNNPpA0JX6vpQyHsmqDy12vEpNtvTQu756aze/679uvrzo1q3PqkUhIZWqzfNRlCf9wkyNHNSbEalYMUi1iEp0Fkq1PEpVGGQpAOzw5m3PzQbrlg8NL0iez+zyf6n68mxPzB2pIw7MZ19bKGqhCT1V7O7zD20Lnxspw/tDC4rAm0KPaBe97ryzfv0mjcdp0eeGe5NCCwGodADrv/0IQ2tcLrw4uyaONCKrv/4sNxtu29W1378kD5zy0ZV6M+AHHCl0MUaDa93v3W/XvHa1Tr5YeQ78kEodLGbrp/Q9LTXxZfxBSXkh7eXLvXQgYb+/IMHdc0XNqha5WMD8rOoUOjGWaDyHtcxpPURJ9szv+PSp64e0zOeu0KnnjEQHCcyOKuUr0jpsvlly1DHp7xniJoOdJYKdXw6WAuX8PbPZJcX909nlyolaf9EoCQ5nuxndja5T7v3HxmjcXYsu7JTGQt3FqtOZj8Gd1XWBdf5rQvvCi4rAlcKXWj3g3V9+YZx3fitDZ0+FPQg2hS60LWfOKQLLhrWxk1kOvLHs6rL7NtT1ze+NK6bvnNCpw8FPYpQ6DLXf3pMz/nNYY1uDH+ORe9p1CWprnPPPXfR69x9990aHR3V1q1bW9oXodBFpqa8br5+XNd9lbYELGxsbKyt9UoVCp2uGMSqLLGZoFoVrArEhmPzTlu+M6lfeeyATjy5Xw0f39ZC+w8Nxxbq+BTq9JQsy34axdYJDdU2Wc++fbwW7hB1aCa7MnFgMlyxODSWXX2oHUgqDL6WnIfZ/Uf2O7A3+/6s2BV+DEa2Zz9zVv3soeA6n33bKRm3/qMkacuWLcH15mvlqqIZDY1d5NYvj+v8C8NlNiAPhEKX2L+3rq23T+s3zst+hwPyQih0iW9/fUJPfeYKDa/kIcPS4hnWJW798oTOfyEfHbD0CIUu8It7a9rxy5rOeUq44QzIy5JVHxqF9Qoor7zOwA+3TOlpz16hvr7WKg3zh1o7fLuvyKdVh/nVhkZoncj7R2hZrPowN9nKYm+fqoefqpO1QCUjMHyaJNUmsrfXfzA5Zld35ndJGtqbff5HdoRrUyP3ZZcF3fbdwXXU4b5GXCl0ga13zOiss8MlOSBPhELJee/1o9un9URCAQUhFErul/fXValIJ27ma80oBqFQcj+6Y1pPOmeQ6eNRGEKh5LbeMaOznsyw7SgOoVByd905oycSCihQqTpEwWo0vH65raZTTws/TKFZoIrUCMxsFTu2UBmz1shuO5kJlColaaaevU5tNtwO46azl80Nn+bq9ndJ6h/LLhUOHKgF91M5OJl5u58u76R+nX9GIWj/3oZWrqxoxQoeJhSHZ1uJ7dhe1wknUnVAsQiFEtvxQF2bTiIUUCxCocR27qjrhE2EAopFKJTYzgdqfHxA4ag+lNiuBxt6zBPaL0eGWv+bOz2FOkDNVw9UGKR4x6vwsdnt1abrmhmvqb5SmV/UClU4Ysti64R6q7m5QoKf97ukSqDIUKmFOzC5WnZnKV/Pc4C/fBEKJTY15bViRW9/k9E3vP76Qz/Wj2+6R74hrTxhRE+76uk6/lfWd/rQli0+PpRYbda33F2622y94W795Oafqz7dUGO2oUPbDuo7r7tNtalw7R9Li1AosVpN6gtPx9gTfvx/7lFtyl5K+4bXjtu3d+iIQCgAMAiFEuvrk2qznT6KpfWE/3ya+oZshcVVnDadc2KHjgiEQon19TvVIi3bveCsi0/X4y96hPoGq6r0V7Tq5NV69sfOU98QbeCdwpkvsaEhp8nJ9kOh4gJ1tzY2WXXhlaqB+l41tH9JlbntVaVnvPkJetrrH6uZ8ZrGh9elJUmf/fexbS3ydklqBN4ODw8r6eb9LinQV0uNSGOw7wusVC3v908IhRLbsLGiXTvLW8/OU99gVX2DVU1M93a1pRvw8aHETjipTzsDcxECS4VQKLETNlW1MzJ8OLAUCIUS23RSVTseIBRQLEKhxDadWOXjAwpHQ2OJrT2+ovHxhiYnG+obys7vYIVBsk3nzes0VQsqSziTV+zYQhWLvkp2CPa5cDj2VQLb6g+vMzuQvU59ODnPc6euPnykgjG7KrsRdOa48MtoYHX2LOGV/ZGObofCi4rAlUKJVSpOm0/u03330A8AxSEUSu6JTx7QXXeWd5BP9B5CoeTOOntAWwkFFIhQKLknnT2oH90+Ld/hmYixfBAKJbf5lKoaDWn7L6lCoBiEQsk55/SkcwZ11x3TnT4ULBOUJJdQKHFbfc8/6+wBbb1jWs9/0cpjPaQFhUqUoRKiJPUHyoXVSM+r/kDpcTAwEOJQNVyBWdmf3eYyORiZVWpl9rLZ2aTs6KvJsc+uPnKcLjDjlAvMUJUYybx11exoeJW9+46+rcBPj1wpdIGnnDukH3xnsue7UaMcCIUu8LCH92nT5j7d/sOpTh8KlgFCoUucf+Gwbr1lotOHgWWAUOgSz7lgWH/zvUlNjC/d15IBiVDoGmuPr+qscwb1/duypzYH8kL1oUQqkSbmivO64EXDuuXGcf3mRcNHFkTaHmPDpLl0xflDpoXWiXZuCiwLVSUkadBlVxNWVLNHql3ZF/5W58xAduv/TGj8NEmNwHhsY+n/ri+5T31rjux3JjDefm1leD+TJ2TvZ9eT1wbXeccXtx1123cfFfzz3BEKXeRpz1qh977jgH55f02bT+GhW47OPffcRf/tXXfdpZGR7JJoDB8fusjQkNNFL1upz1zT4b616AojIyMaHY18HyKAt5su87JXjOjCZ+zUq1+/WqMbyzsiMPJVqUpOVW3ZsmXp97Xke0Cu1q2v6vkv4moBS4dQ6EKX/d4qff3mCT24g8FXkL8l+/hQCeRNYwmH/1qsYBK6yJwDga7L7dybUJ+AuhY358HoxqouvHilPvnRQ3rre8LDejUC+6nKH97T/GMJHdtApJIwG1g2WAnPeTcbGCoudHutL/z+1fCtzxURGsJtsD8J2t19yX06fs3Y4WWTw9nVh/r68LH1V7PPzWnr9gTXGYqctyJwpdClLr1ilb73rUnd81MGYEG+CIUuddyail7zptX6s7fsU71ORynkh1DoYhe9bKUGBp2+cC2NjsgPodDFKhWnt79vnf7yzw9q2309Pmc9CkModLmTT+3Xpa9Zrff88T41GnyMwLEjFHrAS1+xSlOTXrfcMLbwHwMLKNU3GquBkly9jbGoQttqd3shsVRtddi1qgsfVyNYEvWqVqV3fmCdXv3SXXrCWQM649FJmbIe6KiUDLmWbG9+R6d+ZX/3IVYu7XfZZcShSIm3UQk91tm3NxZZrm0Wmm1KCg/vdtxA0gv13nT55lUHDi+rBTpYxTqLrR3I7tX68OFwSTJW/i0CVwo94rQz+vWWd63VH75qj/buZuRntI9Q6CHnvWBYz79opd58xR7NTNO+gPYQCj3m8jes1voNVV31tn1MIIO2EAo9plJxuvKD63T3v83q8//7YKcPB12IUOhBK4Yr+vBfrNcXrz2k226hIoHWLKr6UI20Itc7fIlaCRxbqLW+SNXQaQscWqx5MDRUWyVQsdh0YlUf/eyo/uDluzUz6fXCi4+MwDOg+uF3g4F5e20E3if6I0c35AJfnIq85YT2ExKdWCbYISvci3Skmj3j1tyZGZyrPgwfyPy7xexfko7ry64+bOoPb7c/MFRdUbhS6GGnnTGgT3xhg6792EO64VN8lMDiEAo97uRT+3XNFzfqpuvG9OmPPUTjIxZEKCwDJ5zUp2u+uFHf/uqEPv5+ggFxhMIysX5DVVffuEF3/N2UrnzDbtXpJ4EAQmEZWbM2CYa+fqef/vOMpqY6PwoWyodQWGaGhip6+wfWa/3Gqg4eIBRwtFJ1iCqzYFk28vm81ZdcrOwW6igUWyc4RmPFHx4efrEdomJC5cVo2bGSXRKsBGeoar0kOVwJD1UXGguy7itmmxv6jwxgE5oJKzYWZegYVlXD0/8FS7wF4UoBgEEoADAIBQAGoQDAIBQAGIVXH0IzR0nFzR4VHqotv0qCFJuJquXdh4dqi6wTarGv+vAMUaGKRWhoN6m9ikWoyhA6adVIlWcg0IForpKQuWyB98O56sP65upD4JhjHZhCsz2tDFRfmvfdKVwpADAIBQAGoQDA4BuNQIbd909qeqKuD13y48O3uUB7i4t82zI0OE7o25FSdttFoy65yDp5IhSADCPrsqed7xgnucgIaHnqilBoZ2KX0DBtUjmGamtVsI9D5HkSvJ+ucfjdrX+x7z6xUxY4hli/jFD1IfQOOuvDLfKhKkNsAptGpDIhSe/7yqOPui006Uts8pZQZSLWvyFr2fMef6/6KquC6+SJNgUABqEAwCAUABiEAgCDUABgEAoAjGMuSYaGKev0zFF5C3duitQEWz0HsTJ0cFOxIdzCy1rtEKVYJ53QjFdtfNkm2OkoMkNVw7VekmxHqMQa7NylcLky1okqq0NUMd9QSHClAMAgFAAYhAIAg1AAYBAKAIxSdYgKDdUWG6Yt1Fkq1vLe6rba3l4bE8gELdEQbvO79oY6SFXaaP+OrRNqyQ9VDOqBCoMU7txUWPWhxW7QUnzItYGChiUM4UoBgEEoADAIBQAGoQDAIBQAGIQCAGPJSpJl7igVGr8xNnZjUbNK5Sk2ruPhDlHzy5btPDyBklzVR0q8gXJpqIwZmyEq1FmrHCXJwGjOsfErI6NDF4ErBQAGoQDAIBQAGIQCAINQAGBEqw9zHZRiHZKKEOooJYWPrQyzSrU8hFueHaWkYCWhIn94XsT+eX8U6kQVa8mvBKoMjcj5DFcZAutEWvhDx1bUxG/tVBLis2dRfQBQIoQCAINQAGAQCgAMQgGAQSgAMEo1RmM3ynNcx+CYjlLu5cojHaLm7yd4AK3vJ7JKqOwWK2OGt1Xk/ElHi5UXQ2Jlx6yyMDNEAegYQgGAQSgAMAgFAAahAMBYVClVDOUAAAVTSURBVPWhnQ5JIbEW9naGaitqVqkihnCLnclcZ5vSkdbso85eG5WE0MLgDFUKd2IKzhwVGdqt0x2IYkLnoJ2KRVG4UgBgEAoADEIBgEEoADAIBQBGqfo+dHoCmaKGcMtzYplj7S9RXfAvUpHdhLZRb6fvQ2gfbUyQUtRkMO0o87txmY8NQAcQCgAMQgGAQSgAMAgFAAahAMA45pJkOx2SipBnJy6pmE5UsZJonp2oYiXeRZcoF6OdimCO1ecydJQ6ari7LsCVAgCDUABgEAoADEIBgEEoADBK1SEqpAxDuOWpiKHdopyTS7cX7VBldlNMS34lcDidrWW1rxvfdbvxmAEsIUIBgEEoADAIBQAGoQDAIBQAGEtWkiyq7JfnuI7tdKJqZ1zH4P5zHe8xMq6kIjNEhSy2dNksxzJmrKNWPbe95C/XDmYF4UoBgEEoADAIBQAGoQDA6Iq+D8jfvffVND7udeGL9y7q79tpMvQlGPmoV0xMeFWrk4Xsq/BQyHuYtJAydKKKD692tDyHdlto/8cfX1Er3YzaG1Usv7HIujVgXE7noFKR+vv7c9nWQrhSWKa23LYh8/ZWy6gxeRafi5o6MG+L7oW6gBe9eK8GBp+cy7YWQpsCAINQAGAQCgAMQgGAQSgAMKg+HKM8S6xtdW5qo3W73TJmSKzjVW5yasUvi1bPTZH3nisFAAahAMAgFAAYhAIAg1AAYESrD5W0zbNRUGeUbhzCLSbP+xOqCrTTV6EMFYuQQioZiOJcAzAIBQAGoQDAIBQAGIQCAKMr+j506xBuIUVUJWLKULEIybOSgfZwpQDAIBQAGIQCAINQAGAQCgAMQgGAsaiSZCVSJiqqs1RIpztRxbRaxoyVXkN6rYzZjnZKn53W6rkpcjQ6rhQAGIQCAINQAGAQCgAMQgGAccwdokKVibJWJaT8KxMhRQz7VuaKRUyeU94XVeVYLrhSAGAQCgAMQgGAQSgAMAgFAAahAMBYsjEay1qqlIrrRBVS1AxVIUWVMWPKOhZjO6XSIu5LkWeLKwUABqEAwCAUABiEAgCDUABgFD5DVDcO7SYVU5koYsi3dpWhYlGEslZFisSVAgCDUABgEAoADEIBgEEoADAKrz7EdGN/iRgqFuV9zymqMpLfOSiuKlLeRw1ARxAKAAxCAYBBKAAwCAUABqEAwChVSTKkzJ2oYjo97FtIO2XMmKJKnHkqc7m00zgzAAxCAYBBKAAwCAUABqEAwOiK6kNMmTtRhZS1c1W78qxmdGMlo9dwpQDAIBQAGIQCAINQAGAQCgAMQgGA0fUlyZBu7UQV0mtlzJC8O2v1iiLPClcKAAxCAYBBKAAwCAUABqEAwOjZ6kNMrDIRslwqFiHdWMlAe7hSAGAQCgAMQgGAQSgAMAgFAAahAMBYliXJdnTjWJB5yntGJUqc5cWVAgCDUABgEAoADEIBgBFtaHzWRduLOo6utTyaGZcCZ64VP/mXGY2suruQfUVDwQ38aiEH0c0YPAxFGFl1t0ZHRwvZl/PxabqIc6A3Bd/PaFMAYBAKAAxCAYBBKAAwCAUABqEAwCAUABiEAgCDUABgEAoADEIBgEEoADAIBQAGoQDAIBQAGIQCAINQAGAQCgAMQgGAQSgAMAgFAAahAMAgFAAYhAIAIzpDlJgACVh2uFIAYBAKAAxCAYBBKAAwCAUABqEAwCAUABiEAgCDUABgEAoADEIBgEEoADAIBQAGoQDAIBQAGIQCAINQAGAQCgAMQgGAQSgAMAgFAAahAMAgFAAYhAIAg1AAYBAKAAxCAYBBKAAwCAUABqEAwCAUABj/Dp88dC41f/sQAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 576x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = xg.XGModel.load_model(f'{provider}_openplay_logreg_basic')\n",
    "ax0 = plot_heatmap(model, df, axis=None)\n",
    "ax0.set_title('Logistic regression xG heatmap')\n",
    "\n",
    "model = xg.XGModel.load_model(f'{provider}_openplay_xgboost_basic')\n",
    "ax1 = plot_heatmap(model, df, axis=None);\n",
    "ax1.set_title('Boosted trees xG heatmap')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Headers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "start = np.array(list(itertools.product(np.arange(0, 106, 1), np.arange(0, 69, 1))))\n",
    "df = pd.DataFrame({'start_x': start[:,0], 'start_y': start[:,1], 'bodypart_id': 'head'})\n",
    "df = pd.concat([\n",
    "    fn.startlocation(df), \n",
    "    fn.startpolar(df),\n",
    "    fn.bodypart(df),\n",
    "    fn.goalangle(df),\n",
    "], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = xg.XGModel.load_model(f'{provider}_openplay_logreg_basic')\n",
    "ax0 = plot_heatmap(model, df, axis=None)\n",
    "ax0.set_title('Logistic regression xG heatmap')\n",
    "\n",
    "model = xg.XGModel.load_model(f'{provider}_openplay_xgboost_basic')\n",
    "ax1 = plot_heatmap(model, df, axis=None);\n",
    "ax1.set_title('Boosted trees xG heatmap')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Combining custom models\n",
    "In this notebook we've trained models that can only predict shots from open play. To be able to compute xG values from all shot types, you will have to combine them with other pipelines."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "openplay_model = xg.XGModel.load_model(f'{provider}_openplay_xgboost_basic') # custom pipeline for openplay\n",
    "penalty_model = xg.PenaltyXGModel() # default pipeline for penalties\n",
    "freekick_model = xg.FreekickXGModel() # default pipeline for freekicks\n",
    "\n",
    "model = xg.XGModel()\n",
    "model.model = [openplay_model, penalty_model, freekick_model]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generating features: 100%|██████████| 1446/1446 [06:50<00:00,  3.52it/s]\n",
      "Generating labels: 100%|██████████| 1446/1446 [00:35<00:00, 41.19it/s]\n",
      "Generating features: 100%|██████████| 1446/1446 [04:46<00:00,  5.04it/s]\n",
      "Generating labels: 100%|██████████| 1446/1446 [00:34<00:00, 41.82it/s]\n"
     ]
    }
   ],
   "source": [
    "model.train(api, training_seasons=[('ESP', '1718'), ('ITA', '1718'), ('FRA', '1718'), ('GER', '1718')])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generating labels: 100%|██████████| 380/380 [00:09<00:00, 41.19it/s]\n",
      "Generating features: 100%|██████████| 380/380 [01:48<00:00,  3.52it/s]\n",
      "Generating features: 100%|██████████| 380/380 [00:08<00:00, 42.33it/s]\n",
      "Generating features: 100%|██████████| 380/380 [01:09<00:00,  5.49it/s]\n",
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'max_dev': 29.633825426918392,\n",
       " 'residual_area': 821.0990497947265,\n",
       " 'roc': 0.7871554818983253,\n",
       " 'brier': 0.08370078584210341,\n",
       " 'ece': 0.006895048176804048,\n",
       " 'ace': 0.005767807970400487,\n",
       " 'fig': <Figure size 720x360 with 3 Axes>}"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.validate(api, validation_seasons=[('ENG', '1718')])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "soccer_dataprovider_comparison",
   "language": "python",
   "name": "soccer_dataprovider_comparison"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.2"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}