notebooks/4_feature_engineering.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-05-20T14:35:35.200930Z",
     "start_time": "2018-05-20T14:35:35.194214Z"
    }
   },
   "source": [
    "### Variable Aggregation and Feature Engineering "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:09:14.915352Z",
     "start_time": "2019-05-26T07:09:14.910991Z"
    }
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm_notebook as tqdm\n",
    "from datetime import datetime\n",
    "\n",
    "from constants import DATA_DIR, DF_FEATURES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:09:18.378307Z",
     "start_time": "2019-05-26T07:09:18.371983Z"
    }
   },
   "outputs": [],
   "source": [
    "pd.set_option(\"display.max_columns\", 30)\n",
    "pd.set_option(\"display.max_rows\", 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:09:18.992081Z",
     "start_time": "2019-05-26T07:09:18.986055Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The Cython extension is already loaded. To reload it, use:\n",
      "  %reload_ext Cython\n"
     ]
    }
   ],
   "source": [
    "%load_ext Cython"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataframe for the Baseline Model\n",
    "Mean of past N-1 games for each statistic when predicting N-th game performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:12:08.054517Z",
     "start_time": "2019-05-26T07:12:08.046962Z"
    }
   },
   "outputs": [],
   "source": [
    "def generate_baseline(df):\n",
    "    baseline = {\n",
    "        \"Date\": [],\n",
    "        \"Name\": [],\n",
    "        \"FPTS\": [],\n",
    "        \"PTS\": [],\n",
    "        \"3P\": [],\n",
    "        \"AST\": [],\n",
    "        \"TRB\": [],\n",
    "        \"STL\": [],\n",
    "        \"BLK\": [],\n",
    "        \"TOV\": [],\n",
    "        \"DD\": [],\n",
    "        \"TD\": [],\n",
    "    }\n",
    "\n",
    "    for i in tqdm(range(df.shape[0])):\n",
    "        date = df.loc[i, \"Date\"]\n",
    "        name = df.loc[i, \"Name\"]\n",
    "\n",
    "        df_name = df.loc[df[\"Name\"] == name].reset_index(drop=True)\n",
    "        index = df_name.loc[df_name[\"Date\"] == date].index[0]\n",
    "\n",
    "        # Check if there is past statistics available\n",
    "        # Same baseline as DraftKings official data where season average is taken\n",
    "        if index >= 1:\n",
    "            df_past = df_name[0:index].reset_index(drop=True)\n",
    "\n",
    "            for key in baseline.keys():\n",
    "                if key in [\"Date\", \"Name\", \"FPTS\"]:\n",
    "                    baseline[key].append(df_name.loc[index, key])\n",
    "\n",
    "                else:\n",
    "                    baseline[key].append(df_past[key].mean())\n",
    "\n",
    "    return pd.DataFrame(baseline)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Additional Statistics and Recency Effect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:09:23.837881Z",
     "start_time": "2019-05-26T07:09:23.827468Z"
    }
   },
   "outputs": [],
   "source": [
    "def draw_weights():\n",
    "    weights_dic = {}\n",
    "    weighting = [\"sqrt_0.5\", \"linear_1.0\", \"quad_2.0\"]\n",
    "\n",
    "    for key in weighting:\n",
    "        weights = np.array([np.power(i, float(key[-3:])) for i in range(1, 11)])\n",
    "        weights = weights / weights.sum()\n",
    "        weights_dic[key[:-4]] = weights\n",
    "\n",
    "    sns.set_style(\"darkgrid\")\n",
    "\n",
    "    for key in weights_dic.keys():\n",
    "        plt.plot(weights_dic[key])\n",
    "\n",
    "    plt.xlabel(\"n-th game\", fontsize=12)\n",
    "    plt.ylabel(\"Weight\", fontsize=12)\n",
    "    plt.xticks([i for i in range(0, 10)], [i for i in range(1, 11)])\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:09:24.812260Z",
     "start_time": "2019-05-26T07:09:24.580638Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:28:38.276602Z",
     "start_time": "2019-05-26T07:28:38.271574Z"
    }
   },
   "outputs": [],
   "source": [
    "%%cython\n",
    "def calculate_weighted_mean(list weights, values):\n",
    "    n = len(weights)\n",
    "    weighted_sum = [weights[i] * values[i] for i in range(n)]\n",
    "    weighted_mean = sum(weighted_sum) / sum(weights)\n",
    "    \n",
    "    return weighted_mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:28:15.940546Z",
     "start_time": "2019-05-26T07:28:15.922389Z"
    }
   },
   "outputs": [],
   "source": [
    "def generate_features(df, weighting):\n",
    "    data = {\n",
    "        \"Date\": [],\n",
    "        \"Name\": [],\n",
    "        \"FPTS\": [],\n",
    "        # New feature\n",
    "        \"Rest\": [],\n",
    "        \"Value\": [],\n",
    "        \"FPTS_std\": [],\n",
    "        # Basic 9 Variables\n",
    "        \"PTS\": [],\n",
    "        \"3P\": [],\n",
    "        \"AST\": [],\n",
    "        \"TRB\": [],\n",
    "        \"STL\": [],\n",
    "        \"BLK\": [],\n",
    "        \"TOV\": [],\n",
    "        \"DD\": [],\n",
    "        \"TD\": [],\n",
    "        # Additional Variables from Basketball-Reference.com\n",
    "        \"MP\": [],\n",
    "        \"FT\": [],\n",
    "        \"FTA\": [],\n",
    "        \"FGA\": [],\n",
    "        \"3PA\": [],\n",
    "        \"DRB\": [],\n",
    "        \"ORB\": [],\n",
    "        # Advanced Statistics from Basketball-Reference.com\n",
    "        \"USG_perc\": [],\n",
    "        \"DRtg\": [],\n",
    "        \"ORtg\": [],\n",
    "        \"AST_perc\": [],\n",
    "        \"DRB_perc\": [],\n",
    "        \"ORB_perc\": [],\n",
    "        \"BLK_perc\": [],\n",
    "        \"TOV_perc\": [],\n",
    "        \"STL_perc\": [],\n",
    "        \"eFG_perc\": [],\n",
    "        \"FG_perc\": [],\n",
    "        \"3P_perc\": [],\n",
    "        \"FT_perc\": [],\n",
    "    }\n",
    "\n",
    "    for i in tqdm(range(df.shape[0])):\n",
    "        date = df.loc[i, \"Date\"]\n",
    "        name = df.loc[i, \"Name\"]\n",
    "\n",
    "        df_name = df.loc[df[\"Name\"] == name].reset_index(drop=True)\n",
    "        index = df_name.loc[df_name[\"Date\"] == date].index[0]\n",
    "\n",
    "        # Generate features from the past 10 games\n",
    "        if index >= 10:\n",
    "            df_past = df_name[index - 10 : index].reset_index(drop=True)\n",
    "\n",
    "            # Consider the number of days between the current game and the previous game\n",
    "            current = datetime.strptime(str(df_name.loc[index, \"Date\"]), \"%Y%m%d\")\n",
    "            previous = datetime.strptime(\n",
    "                str(df_past.loc[df_past.shape[0] - 1, \"Date\"]), \"%Y%m%d\"\n",
    "            )\n",
    "            rest = current - previous\n",
    "\n",
    "            data[\"Rest\"].append(rest.days)\n",
    "\n",
    "            # Weights higehr towards the most recent game\n",
    "            if weighting == \"linear\":\n",
    "                weights = [i for i in range(1, 11)]\n",
    "\n",
    "            elif weighting == \"quad\":\n",
    "                weights = [i**2 for i in range(1, 11)]\n",
    "\n",
    "            elif weighting == \"sqrt\":\n",
    "                weights = [i ** (1 / 2) for i in range(1, 11)]\n",
    "\n",
    "            for key in data.keys():\n",
    "                if key in [\"Date\", \"Name\", \"FPTS\"]:\n",
    "                    data[key].append(df_name.loc[index, key])\n",
    "                elif key == \"FPTS_std\":\n",
    "                    data[key].append(df_past[\"FPTS\"].std())\n",
    "                elif key != \"Rest\":\n",
    "                    weighted_mean = calculate_weighted_mean(weights, df_past[key])\n",
    "                    data[key].append(weighted_mean)\n",
    "\n",
    "    return pd.DataFrame(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Add Roster Availability "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:28:14.132303Z",
     "start_time": "2019-05-26T07:28:14.123208Z"
    }
   },
   "outputs": [],
   "source": [
    "def add_roster_info(df):\n",
    "    df[\"Rota_All\"] = [0 for i in range(df.shape[0])]\n",
    "    df[\"Rota_Pos\"] = [0 for i in range(df.shape[0])]\n",
    "\n",
    "    for date in tqdm(list(set(df[\"Date\"]))):\n",
    "        for team in list(set(df[\"Team\"])):\n",
    "            df_rota = df.loc[(df[\"Date\"] == date) & (df[\"Team\"] == team)]\n",
    "\n",
    "            if df_rota.shape[0] != 0:\n",
    "                rota_all = df_rota.shape[0]\n",
    "\n",
    "                for pos in [\"PG\", \"SG\", \"F\", \"C\"]:\n",
    "                    df_pos = df_rota.loc[df_rota[pos] == 1]\n",
    "                    rota_pos = df_pos.shape[0]\n",
    "\n",
    "                    df.loc[\n",
    "                        (df[\"Date\"] == date) & (df[\"Team\"] == team) & (df[pos] == 1),\n",
    "                        \"Rota_All\",\n",
    "                    ] = rota_all\n",
    "                    df.loc[\n",
    "                        (df[\"Date\"] == date) & (df[\"Team\"] == team) & (df[pos] == 1),\n",
    "                        \"Rota_Pos\",\n",
    "                    ] = rota_pos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:11:14.219213Z",
     "start_time": "2019-05-26T07:11:14.212928Z"
    }
   },
   "outputs": [],
   "source": [
    "seasons = [\"2014-15\", \"2015-16\", \"2016-17\", \"2017-18\", \"2018-19\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T07:28:01.975421Z",
     "start_time": "2019-05-26T07:12:51.553862Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "11b80ff874654255ad676e65bb78d37e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "eded8caa98714204ad0e29209c693d0a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=25332), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1ee90d6e8f2a4ed2ac9d5c522e8e4b61",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=30595), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "56e40ea3b6b84c63bb9f6ab86622cf71",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29538), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "49a1ca311cf84e71a72d8c6e2b17981d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29271), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1d57d63efcff464b8036d20ea502edc2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=28911), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Takes 15-20 mins\n",
    "for season in tqdm(seasons):\n",
    "    df = pd.read_csv(\n",
    "        os.path.join(DATA_DIR, \"Dataframes\", \"Merged\", \"df_{}.csv\".format(season))\n",
    "    )\n",
    "    df_baseline = generate_baseline(df)\n",
    "    df_baseline = df_baseline.loc[\n",
    "        :,\n",
    "        [\n",
    "            \"Date\",\n",
    "            \"Name\",\n",
    "            \"FPTS\",\n",
    "            \"PTS\",\n",
    "            \"3P\",\n",
    "            \"AST\",\n",
    "            \"TRB\",\n",
    "            \"STL\",\n",
    "            \"BLK\",\n",
    "            \"TOV\",\n",
    "            \"DD\",\n",
    "            \"TD\",\n",
    "        ],\n",
    "    ]\n",
    "    df_baseline.to_csv(\n",
    "        os.path.join(\n",
    "            DATA_DIR, \"Dataframes\", \"Modelling\", \"Baseline\", \"{}.csv\".format(season)\n",
    "        ),\n",
    "        index=False,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-26T09:27:59.904540Z",
     "start_time": "2019-05-26T07:40:43.192362Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "96d1ee84d9914ca481062847355de493",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=3), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ec8fdea87a6145cc912edd02c616764e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7cdcd8f3127d4962ba659fc0510037c7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=25332), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b0f6d0c9e9ce44029bb18ba91c2562ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=144), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9014cd126b8c47e2924c6f94046afb21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=30595), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "40bcc8be53e3457eb8566569a6144cea",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=144), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bc7d89c22d1c422bb36a129d162dbb86",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29538), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d969d0e1af524ea8bd834686ad6ee785",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=145), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "729db4793ece4b6980b8abbd333e8b9c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29271), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2e17d7800b764fbc8be0c15404aa4187",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=150), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "63aaee1fb3b54015873c7f3cea545c82",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=28911), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a909871abb6841a68efa038abd67accc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=149), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f1cd13573b7242829f59526dfc67f52c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "40216abadf3d4708a91c34197132d5cc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=25332), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5605b7c06e614edb9d9baeab52a23295",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=144), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "119ad940992843438f45d42b016fdeda",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=30595), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "191cd56b2cd44460ac7bd4159a323ec8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=144), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aea1ca7da80f4ee48558f2060e7e4988",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29538), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "afaf581cf5084a598a15e450d8ff735b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=145), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "489e910db39145a4b1c4c826e1f396cd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29271), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "933957f3a1bf4a1bbab00a5e2360f90e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=150), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "39e7509c4fb747f4a8f2acdfb99bad5d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=28911), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "38716f245ed4416b9c7cde19c6dedafa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=149), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bcdbb9c8abb04dd9b3eec3b82ee0de20",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "da99bc1c4ee949eabb33ea4015e4e48c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=25332), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "17fb1317235e40e9928a0414657e8f21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=144), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5cbde29d1d294ff08a1cd143bf7ec25d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=30595), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cccbe81e819747ab8d775dab21f6ade4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=144), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9eb972338c5545d082a01161197e66bc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29538), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a0d9f79b8acb4a429363328bb8e1f3d8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=145), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c255749abbaf4756b49f9ae22443bcae",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=29271), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3ef46b6fd4e04ddd964c61536b3b8824",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=150), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3a0e0bf524d640a59549873aef616ee9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=28911), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f64b284663164ae48a14cb5e07003a1f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=149), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "weighting_types = [\"sqrt\", \"linear\", \"quad\"]\n",
    "\n",
    "# Takes ~ 2 hrs in total\n",
    "# TODO: Optimize\n",
    "for weighting in tqdm(weighting_types):\n",
    "    for season in tqdm(seasons):\n",
    "        df = pd.read_csv(\n",
    "            os.path.join(DATA_DIR, \"Dataframes\", \"Merged\", \"df_{}.csv\".format(season))\n",
    "        )\n",
    "        df_features = generate_features(df, weighting)\n",
    "        # Add Starter, Listed Position and Team\n",
    "        df_features = pd.merge(\n",
    "            df.loc[\n",
    "                :,\n",
    "                [\n",
    "                    \"Date\",\n",
    "                    \"Name\",\n",
    "                    \"Salary\",\n",
    "                    \"Team\",\n",
    "                    \"Starter\",\n",
    "                    \"Home\",\n",
    "                    \"Pos\",\n",
    "                    \"PG\",\n",
    "                    \"SG\",\n",
    "                    \"F\",\n",
    "                    \"C\",\n",
    "                ],\n",
    "            ],\n",
    "            df_features,\n",
    "            on=[\"Date\", \"Name\"],\n",
    "            how=\"inner\",\n",
    "        )\n",
    "        # Add roster information\n",
    "        add_roster_info(df_features)\n",
    "        df_features = df_features.loc[:, DF_FEATURES]\n",
    "        df_features.to_csv(\n",
    "            os.path.join(\n",
    "                DATA_DIR,\n",
    "                \"Dataframes\",\n",
    "                \"Modelling\",\n",
    "                \"Features\",\n",
    "                weighting,\n",
    "                \"{}.csv\".format(season),\n",
    "            ),\n",
    "            index=False,\n",
    "        )"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}