notebooks/01_Preprocesamiento.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"6\"></a> <br>\n",
    "## PREPROCESAMIENTO DEL DATASET Y CREACIÓN DE NUEVAS VARIABLES"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Indice:\n",
    "1. [Limpieza de valores nulos](#1)  \n",
    "2. [Creación de nuevas columnas](#2)  \n",
    "    2.1 [Edición columna Domains](#3)  \n",
    "    2.2 [Edición columna Mechanics](#4)\n",
    "3. [Enlace a EDA](#5)  "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"1\"></a> <br>\n",
    "### 1. Limpieza de valores nulos."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Para empezar la limpieza de datos empezaros primero cargando librerias e ir limpiando columna a columna valores nulos, datos incorrectos..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "np.random.seed(22)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"../data/raw/bgg_dataset.csv\",sep=\";\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>ID</th>\n",
       "      <th>Name</th>\n",
       "      <th>Year Published</th>\n",
       "      <th>Min Players</th>\n",
       "      <th>Max Players</th>\n",
       "      <th>Play Time</th>\n",
       "      <th>Min Age</th>\n",
       "      <th>Users Rated</th>\n",
       "      <th>Rating Average</th>\n",
       "      <th>BGG Rank</th>\n",
       "      <th>Complexity Average</th>\n",
       "      <th>Owned Users</th>\n",
       "      <th>Mechanics</th>\n",
       "      <th>Domains</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>174430.0</td>\n",
       "      <td>Gloomhaven</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>120</td>\n",
       "      <td>14</td>\n",
       "      <td>42055</td>\n",
       "      <td>8,79</td>\n",
       "      <td>1</td>\n",
       "      <td>3,86</td>\n",
       "      <td>68323.0</td>\n",
       "      <td>Action Queue, Action Retrieval, Campaign / Bat...</td>\n",
       "      <td>Strategy Games, Thematic Games</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>161936.0</td>\n",
       "      <td>Pandemic Legacy: Season 1</td>\n",
       "      <td>2015.0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>60</td>\n",
       "      <td>13</td>\n",
       "      <td>41643</td>\n",
       "      <td>8,61</td>\n",
       "      <td>2</td>\n",
       "      <td>2,84</td>\n",
       "      <td>65294.0</td>\n",
       "      <td>Action Points, Cooperative Game, Hand Manageme...</td>\n",
       "      <td>Strategy Games, Thematic Games</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>224517.0</td>\n",
       "      <td>Brass: Birmingham</td>\n",
       "      <td>2018.0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>120</td>\n",
       "      <td>14</td>\n",
       "      <td>19217</td>\n",
       "      <td>8,66</td>\n",
       "      <td>3</td>\n",
       "      <td>3,91</td>\n",
       "      <td>28785.0</td>\n",
       "      <td>Hand Management, Income, Loans, Market, Networ...</td>\n",
       "      <td>Strategy Games</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>167791.0</td>\n",
       "      <td>Terraforming Mars</td>\n",
       "      <td>2016.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>120</td>\n",
       "      <td>12</td>\n",
       "      <td>64864</td>\n",
       "      <td>8,43</td>\n",
       "      <td>4</td>\n",
       "      <td>3,24</td>\n",
       "      <td>87099.0</td>\n",
       "      <td>Card Drafting, Drafting, End Game Bonuses, Han...</td>\n",
       "      <td>Strategy Games</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>233078.0</td>\n",
       "      <td>Twilight Imperium: Fourth Edition</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>480</td>\n",
       "      <td>14</td>\n",
       "      <td>13468</td>\n",
       "      <td>8,70</td>\n",
       "      <td>5</td>\n",
       "      <td>4,22</td>\n",
       "      <td>16831.0</td>\n",
       "      <td>Action Drafting, Area Majority / Influence, Ar...</td>\n",
       "      <td>Strategy Games, Thematic Games</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ID                               Name  Year Published  Min Players  \\\n",
       "0  174430.0                         Gloomhaven          2017.0            1   \n",
       "1  161936.0          Pandemic Legacy: Season 1          2015.0            2   \n",
       "2  224517.0                  Brass: Birmingham          2018.0            2   \n",
       "3  167791.0                  Terraforming Mars          2016.0            1   \n",
       "4  233078.0  Twilight Imperium: Fourth Edition          2017.0            3   \n",
       "\n",
       "   Max Players  Play Time  Min Age  Users Rated Rating Average  BGG Rank  \\\n",
       "0            4        120       14        42055           8,79         1   \n",
       "1            4         60       13        41643           8,61         2   \n",
       "2            4        120       14        19217           8,66         3   \n",
       "3            5        120       12        64864           8,43         4   \n",
       "4            6        480       14        13468           8,70         5   \n",
       "\n",
       "  Complexity Average  Owned Users  \\\n",
       "0               3,86      68323.0   \n",
       "1               2,84      65294.0   \n",
       "2               3,91      28785.0   \n",
       "3               3,24      87099.0   \n",
       "4               4,22      16831.0   \n",
       "\n",
       "                                           Mechanics  \\\n",
       "0  Action Queue, Action Retrieval, Campaign / Bat...   \n",
       "1  Action Points, Cooperative Game, Hand Manageme...   \n",
       "2  Hand Management, Income, Loans, Market, Networ...   \n",
       "3  Card Drafting, Drafting, End Game Bonuses, Han...   \n",
       "4  Action Drafting, Area Majority / Influence, Ar...   \n",
       "\n",
       "                          Domains  \n",
       "0  Strategy Games, Thematic Games  \n",
       "1  Strategy Games, Thematic Games  \n",
       "2                  Strategy Games  \n",
       "3                  Strategy Games  \n",
       "4  Strategy Games, Thematic Games  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 20343 entries, 0 to 20342\n",
      "Data columns (total 14 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   ID                  20327 non-null  float64\n",
      " 1   Name                20343 non-null  object \n",
      " 2   Year Published      20342 non-null  float64\n",
      " 3   Min Players         20343 non-null  int64  \n",
      " 4   Max Players         20343 non-null  int64  \n",
      " 5   Play Time           20343 non-null  int64  \n",
      " 6   Min Age             20343 non-null  int64  \n",
      " 7   Users Rated         20343 non-null  int64  \n",
      " 8   Rating Average      20343 non-null  object \n",
      " 9   BGG Rank            20343 non-null  int64  \n",
      " 10  Complexity Average  20343 non-null  object \n",
      " 11  Owned Users         20320 non-null  float64\n",
      " 12  Mechanics           18745 non-null  object \n",
      " 13  Domains             10184 non-null  object \n",
      "dtypes: float64(3), int64(6), object(5)\n",
      "memory usage: 2.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Queremos limpiar nulls y datos que no concuerdan como las comas y los puntos."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"Complexity Average\"] = df[\"Complexity Average\"].str.replace(\",\",\".\").astype(float) # Cambiamos las comas por puntos y pasamos a float\n",
    "df[\"Rating Average\"] = df[\"Rating Average\"].str.replace(\",\",\".\").astype(float) # Cambiamos las comas por puntos y pasamos a float\n",
    "media_valores_año = int(np.mean(df.loc[df[\"Year Published\"] >= 0, \"Year Published\"])) # Hacemos la media de los años para dar un valor concreto a los años\n",
    "df.loc[df[\"Year Published\"] < 1, \"Year Published\"] = media_valores_año # Miramos que todos los años menores de 1 sean\n",
    "df.loc[df[\"Year Published\"].isna(), \"Year Published\"] = media_valores_año\n",
    "df[\"Year Published\"] = df[\"Year Published\"].astype(int) # Pasamos a int todos los años\n",
    "df.loc[(df[\"Play Time\"] >= 0) & (df[\"Play Time\"] <= 4), \"Play Time\"] = 5 # Como es bastante complicado jugar a un juego que se tar 0,1,2,3 o 4 min los convierto todos a 5\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Quitamos de la fila de Domains los null como no definidos."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.loc[df[\"Domains\"].isnull(), \"Domains\"] = \"Not Defined\" # Definimos como \"Not Defined\" los valores null del dataset"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En Mechanics ponemos los nulls igual que en domains como \"Not Defined\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.loc[df[\"Mechanics\"].isnull(), \"Mechanics\"] = \"Not Defined\" # Definimos como \"Not Defined\" los valores null del dataset"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En Owned Users ponemos los nulls igual que en domains como \"Not Defined\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "media_valores_owned = int(np.mean(df.loc[df[\"Owned Users\"] >= 0, \"Owned Users\"]))\n",
    "df.loc[df[\"Owned Users\"].isna(), \"Owned Users\"] = media_valores_owned # Definimos los valores null como media de owned por juego.\n",
    "df[\"Owned Users\"] = df[\"Owned Users\"].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        8.79\n",
       "1        8.61\n",
       "2        8.66\n",
       "3        8.43\n",
       "4        8.70\n",
       "         ... \n",
       "20338    2.28\n",
       "20339    2.85\n",
       "20340    3.18\n",
       "20341    2.86\n",
       "20342    2.68\n",
       "Name: Rating Average, Length: 20343, dtype: float64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"Rating Average\"]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En ID tendremos que asignar valores que no estén repetidos en las columnas. Quitamos los valores nulos y sustituimos por valores aleatorios.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "valores_nulos = df[\"ID\"].isnull() # Vemos los valores nulos\n",
    "valores_no_nulos = df[\"ID\"].dropna().unique() # Sacamos los valores que no son nulos\n",
    "valores_aleatorios = np.random.choice(valores_no_nulos, size=valores_nulos.sum(), replace=False) # Hacemos valores aleatorios en una variable\n",
    "df.loc[valores_nulos, \"ID\"] = valores_aleatorios # Cambiamos valores nulos y sustituimos por valores aleatorios \n",
    "df[\"ID\"] = df[\"ID\"].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 20343 entries, 0 to 20342\n",
      "Data columns (total 14 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   ID                  20343 non-null  int32  \n",
      " 1   Name                20343 non-null  object \n",
      " 2   Year Published      20343 non-null  int32  \n",
      " 3   Min Players         20343 non-null  int64  \n",
      " 4   Max Players         20343 non-null  int64  \n",
      " 5   Play Time           20343 non-null  int64  \n",
      " 6   Min Age             20343 non-null  int64  \n",
      " 7   Users Rated         20343 non-null  int64  \n",
      " 8   Rating Average      20343 non-null  float64\n",
      " 9   BGG Rank            20343 non-null  int64  \n",
      " 10  Complexity Average  20343 non-null  float64\n",
      " 11  Owned Users         20343 non-null  int32  \n",
      " 12  Mechanics           20343 non-null  object \n",
      " 13  Domains             20343 non-null  object \n",
      "dtypes: float64(2), int32(3), int64(6), object(3)\n",
      "memory usage: 1.9+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Volver al índice](#6)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"2\"></a> <br>\n",
    "### 2. Creación de nuevas columnas"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"3\"></a> <br>\n",
    "> #### 2.1 Edición de columna Domains"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ahora toca limpiar las columnas mechanics y domains, quizas las que más trabajo llevan. Primero empezaré con Domains"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.join(df['Domains'].str.split(expand=True))\n",
    "df = pd.get_dummies(df, columns=[0,1,2,3,4,5])\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ahora vamos asociando cada columna creada con el get dummies y poniendo nombre a cada tipo de juego dejandolo todo en 0 y 1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"Children\"] = df[[\"0_Children's\", \"2_Children's\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Customizable\"] = df[[\"2_Customizable\", \"0_Customizable\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Family\"] = df[[\"2_Family\", \"0_Family\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Party\"] = df[[\"0_Party\", \"2_Party\",\"4_Party\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Strategy\"] = df[[\"0_Strategy\", \"2_Strategy\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Thematic\"] = df[[\"0_Thematic\", \"2_Thematic\",\"4_Thematic\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Wargames\"] = df[[\"0_Wargames\", \"2_Wargames\",\"4_Wargames\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Domain_Not Defined\"] = df[[\"0_Not\",\"1_Defined\"]].any(axis='columns').astype(np.uint8)\n",
    "df = df.rename(columns={\"0_Abstract\":\"Abstract\"})\n",
    "df[\"Abstract\"].astype(np.uint8)\n",
    "df.drop([\"0_Children's\", '0_Customizable', '0_Family', '0_Not',\n",
    "       '0_Party', '0_Strategy', '0_Thematic', '0_Wargames', '1_Defined',\n",
    "       '1_Games', '1_Games,', \"2_Children's\", '2_Customizable', '2_Family',\n",
    "       '2_Party', '2_Strategy', '2_Thematic', '2_Wargames', '3_Games',\n",
    "       '3_Games,', '4_Party', '4_Thematic', '4_Wargames', '5_Games'],inplace=True,axis=1)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"4\"></a> <br>\n",
    "> #### 2.2 Edición de columna Mechanics"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ahora toca hacer algo parecido con la columna Mechanics, al haber bastante más tipos de registros de mecanicas de juego.  \n",
    "Primero separaremos por las comas la columna de Mechanics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_descriptions = df['Mechanics'].str.get_dummies(', ')\n",
    "\n",
    "df = pd.concat([df, df_descriptions], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ID', 'Name', 'Year Published', 'Min Players', 'Max Players',\n",
       "       'Play Time', 'Min Age', 'Users Rated', 'Rating Average', 'BGG Rank',\n",
       "       ...\n",
       "       'Turn Order: Role Order', 'Turn Order: Stat-Based',\n",
       "       'Variable Phase Order', 'Variable Player Powers', 'Variable Set-up',\n",
       "       'Victory Points as a Resource', 'Voting', 'Worker Placement',\n",
       "       'Worker Placement with Dice Workers', 'Zone of Control'],\n",
       "      dtype='object', length=206)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ahora toca procesar informacion y unir muchas mecanicas de juegos para dejarlo lo mas generalizado posible. (Dejo un [archivo de texto](https://github.com/Kuja182/Prediccion-sobre-juegos-de-mesa/blob/main/docs/Mechanics_selec.txt) donde ordené de manera preliminar las mechanicas para generalizar al máximo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"Mech_Acting\"] = df[[\"Communication Limits\",\"Alliances\",\"Acting\",\"Player Judge\",\"Role Playing\",\"Roles with Asymmetric Information\",\"Storytelling\", \"Traitor Game\",\"Catch the Leader\",\"Follow\",\"Force Commitment\",\"Hidden Roles\",\"Negotiation\",\"Voting\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_Action\"] = df[[\"Static Capture\",\"Sudden Death Ending\",\"Simultaneous Action Selection\",\"Matching\",\"Cube Tower\",\"Bias\",\"Real-Time\",\"Push Your Luck\",\"Action Drafting\", \"Action Points\",\"Action Queue\",\"Action Retrieval\",\"Action Timer\",\"Action/Event\",\"Critical Hits and Failures\",\"Hot Potato\",\"I Cut You Choose\",\"Interrupts\",\"Kill Steal\",\"Ladder Climbing\",\"Line Drawing\",\"Paper-and-Pencil\",\"Race\",\"Singing\",\"Take That\",\"Three Dimensional Movement\",\"Track Movement\",\"Measurement Movement\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_tokens\"] = df[[\"Physical Removal\",\"Advantage Token\", \"Passed Action Token\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_construcc_farm\"] = df[[\"Tech Trees / Tech Tracks\",\"Resource to Move\",\"Automatic Resource Growth\",\"Random Production\",\"Mancala\",\"Network and Route Building\",\"Ownership\",\"Pattern Building\",\"Pick-up and Deliver\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_roll_thng\"] = df[[\"Variable Set-up\",\"Worker Placement with Dice Workers\",\"Roll / Spin and Move\",\"Flicking\",\"Re-rolling and Locking\",\"Rondel\",\"Dice Rolling\",\"Different Dice Movement\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_cards\"] = df[[\"Trick-taking\",\"Melding and Splaying\",\"Drafting\",\"Set Collection\",\"Move Through Deck\",\"Hand Management\",\"Deck Construction\",\"Deck Bag and Pool Building\",\"Campaign / Battle Card Driven\",\"Card Drafting\",\"Card Play Conflict Resolution\",\"Chaining\",\"Command Cards\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_role_camp\"] = df[[\"Variable Player Powers\",\"Worker Placement\",\"Stat Check Resolution\",\"Simulation\",\"Scenario / Mission / Campaign Game\",\"Ratio / Combat Results Table\",\"Narrative Choice / Paragraph\",\"Multiple Maps\",\"Legacy Game\",\"Finale Ending\",\"Events\",\"End Game Bonuses\",\"Enclosure\",\"Elapsed Real Time Ending\",\"Different Worker Types\",\"Die Icon Resolution\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_board\"] = df[[\"Tile Placement\",\"Square Grid\",\"Slide/Push\",\"Line of Sight\",\"Zone of Control\",\"Secret Unit Deployment\",\"Point to Point Movement\",\"Pieces as Map\",\"Moving Multiple Units\",\"Movement Template\",\"Movement Points\",\"Modular Board\",\"Minimap Resolution\",\"Map Reduction\",\"Map Deformation\",\"Map Addition\",\"Layering\",\"Impulse Movement\",\"Hidden Movement\",\"Hexagon Grid\",\"Grid Movement\",\"Grid Coverage\",\"Crayon Rail System\",\"Connections\",\"Area Movement\",\"Area Majority / Influence\",\"Area-Impulse\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_money\"] = df[[\"Trading\",\"Stock Holding\",\"Stacking and Balancing\",\"Selection Order Bid\",\"Predictive Bid\",\"Order Counters\",\"Multiple-Lot Auction\",\"Market\",\"Loans\",\"Investment\",\"Increase Value of Unchosen Resources\",\"Income\",\"Delayed Purchase\",\"Contracts\",\"Closed Economy Auction\",\"Constrained Bidding\",\"Commodity Speculation\",\"Bribery\",\"Bingo\",\"Betting and Bluffing\",\"Auction/Bidding\",\"Auction: Dexterity\",\"Auction: Dutch\",\"Auction: Dutch Priority\",\"Auction: English\",\"Auction: Fixed Placement\",\"Auction: Once Around\",\"Auction: Sealed Bid\",\"Auction: Turn Order Until Pass\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_score\"] = df[[\"Victory Points as a Resource\",\"Player Elimination\",\"Highest-Lowest Scoring\",\"Hidden Victory Points\",\"Score-and-Reset Game\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_turnbased\"] = df[[\"Variable Phase Order\",\"Turn Order: Stat-Based\",\"Turn Order: Role Order\",\"Turn Order: Random\",\"Turn Order: Progressive\",\"Turn Order: Pass Order\",\"Turn Order: Claim Action\",\"Turn Order: Auction\",\"Relative Movement\",\"Programmed Movement\",\"Pattern Recognition\",\"Pattern Movement\",\"Lose a Turn\",\"Chit-Pull System\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_team\"] = df[[\"Tug of War\",\"Prisoner's Dilemma\",\"Team-Based Game\",\"Semi-Cooperative Game\",\"Cooperative Game\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_skill\"] = df[[\"Time Track\",\"Targeted Clues\",\"Induction\",\"Deduction\",\"Speed Matching\",\"Rock-Paper-Scissors\",\"Once-Per-Game Abilities\",\"Memory\"]].any(axis='columns').astype(np.uint8)\n",
    "df[\"Mech_solo\"] = df[[\"King of the Hill\",\"Single Loser Game\",\"Solo / Solitaire Game\"]].any(axis='columns').astype(np.uint8)\n",
    "df = df.rename(columns={\"Not Defined\":\"Mech Not Defined\"})\n",
    "df.drop([\"King of the Hill\",\"Communication Limits\",\"Alliances\",\"Acting\",\"Player Judge\",\"Role Playing\",\"Roles with Asymmetric Information\",\"Storytelling\", \"Traitor Game\",\"Catch the Leader\",\"Follow\",\"Force Commitment\",\"Hidden Roles\",\"Negotiation\",\"Voting\",\"Static Capture\",\"Sudden Death Ending\",\"Simultaneous Action Selection\",\"Matching\",\"Cube Tower\",\"Bias\",\"Real-Time\",\"Push Your Luck\",\"Action Drafting\", \"Action Points\",\"Action Queue\",\"Action Retrieval\",\"Action Timer\",\"Action/Event\",\"Critical Hits and Failures\",\"Hot Potato\",\"I Cut You Choose\",\"Interrupts\",\"Kill Steal\",\"Ladder Climbing\",\"Line Drawing\",\"Paper-and-Pencil\",\"Race\",\"Singing\",\"Take That\",\"Three Dimensional Movement\",\"Track Movement\",\"Measurement Movement\",\"Physical Removal\",\"Advantage Token\", \"Passed Action Token\",\"Tech Trees / Tech Tracks\",\"Resource to Move\",\"Automatic Resource Growth\",\"Random Production\",\"Mancala\",\"Network and Route Building\",\"Ownership\",\"Pattern Building\",\"Pick-up and Deliver\",\"Variable Set-up\",\"Worker Placement with Dice Workers\",\"Roll / Spin and Move\",\"Flicking\",\"Re-rolling and Locking\",\"Rondel\",\"Dice Rolling\",\"Different Dice Movement\",\"Trick-taking\",\"Melding and Splaying\",\"Drafting\",\"Set Collection\",\"Move Through Deck\",\"Hand Management\",\"Deck Construction\",\"Deck Bag and Pool Building\",\"Campaign / Battle Card Driven\",\"Card Drafting\",\"Card Play Conflict Resolution\",\"Chaining\",\"Command Cards\",\"Variable Player Powers\",\"Worker Placement\",\"Stat Check Resolution\",\"Simulation\",\"Scenario / Mission / Campaign Game\",\"Ratio / Combat Results Table\",\"Narrative Choice / Paragraph\",\"Multiple Maps\",\"Legacy Game\",\"Finale Ending\",\"Events\",\"End Game Bonuses\",\"Enclosure\",\"Elapsed Real Time Ending\",\"Different Worker Types\",\"Die Icon Resolution\",\"Tile Placement\",\"Square Grid\",\"Slide/Push\",\"Line of Sight\",\"Zone of Control\",\"Secret Unit Deployment\",\"Point to Point Movement\",\"Pieces as Map\",\"Moving Multiple Units\",\"Movement Template\",\"Movement Points\",\"Modular Board\",\"Minimap Resolution\",\"Map Reduction\",\"Map Deformation\",\"Map Addition\",\"Layering\",\"Impulse Movement\",\"Hidden Movement\",\"Hexagon Grid\",\"Grid Movement\",\"Grid Coverage\",\"Crayon Rail System\",\"Connections\",\"Area Movement\",\"Area Majority / Influence\",\"Area-Impulse\",\"Trading\",\"Stock Holding\",\"Stacking and Balancing\",\"Selection Order Bid\",\"Predictive Bid\",\"Order Counters\",\"Multiple-Lot Auction\",\"Market\",\"Loans\",\"Investment\",\"Increase Value of Unchosen Resources\",\"Income\",\"Delayed Purchase\",\"Contracts\",\"Closed Economy Auction\",\"Constrained Bidding\",\"Commodity Speculation\",\"Bribery\",\"Bingo\",\"Betting and Bluffing\",\"Auction/Bidding\",\"Auction: Dexterity\",\"Auction: Dutch\",\"Auction: Dutch Priority\",\"Auction: English\",\"Auction: Fixed Placement\",\"Auction: Once Around\",\"Auction: Sealed Bid\",\"Auction: Turn Order Until Pass\",\"Victory Points as a Resource\",\"Player Elimination\",\"Highest-Lowest Scoring\",\"Hidden Victory Points\",\"Score-and-Reset Game\",\"Variable Phase Order\",\"Turn Order: Stat-Based\",\"Turn Order: Role Order\",\"Turn Order: Random\",\"Turn Order: Progressive\",\"Turn Order: Pass Order\",\"Turn Order: Claim Action\",\"Turn Order: Auction\",\"Relative Movement\",\"Programmed Movement\",\"Pattern Recognition\",\"Pattern Movement\",\"Lose a Turn\",\"Chit-Pull System\",\"Tug of War\",\"Prisoner's Dilemma\",\"Team-Based Game\",\"Semi-Cooperative Game\",\"Cooperative Game\",\"Time Track\",\"Targeted Clues\",\"Induction\",\"Deduction\",\"Speed Matching\",\"Rock-Paper-Scissors\",\"Once-Per-Game Abilities\",\"Memory\",\"Single Loser Game\",\"Solo / Solitaire Game\"],inplace=True,axis=1)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Volver al índice](#6)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"5\"></a> <br>\n",
    "### 3. Enlace a EDA  \n",
    "Una vez ya tenemos todos los features ordenados, sin valores nulos y lo mejor clasificado posible vamos a hacer el [analisis de datos](https://github.com/Kuja182/Prediccion-sobre-juegos-de-mesa/blob/main/notebooks/02_EDA.ipynb) y el [modelo de predicción](https://github.com/Kuja182/Prediccion-sobre-juegos-de-mesa/blob/main/notebooks/03_Entrenamiento_Modelo.ipynb)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Volver al índice](#6)"
   ]
  }
 ],
 "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": "3.7.4"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}