WyScout Dataset - Model Selection ( Train, Test) at Level 2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Author**: Fabrizio Lucero Fernández. https://www.linkedin.com/in/fabrizio-lucero/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font size=\"4.5\">**Data Analytics in the Sports World: Assesing 2017/2018 English Premier League Data through Machine Learning**</font>"
   ]
  },
  {
   "attachments": {
    "1200px-Memorial_University_of_Newfoundland_Logo.svg.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![1200px-Memorial_University_of_Newfoundland_Logo.svg.png](attachment:1200px-Memorial_University_of_Newfoundland_Logo.svg.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Libraries**: Appart from the usual libraries we have been using throughout the project now we added the confusion matrix from ```sk_learn.metrics```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "importing Jupyter notebook from definitions.ipynb\n"
     ]
    }
   ],
   "source": [
    "import import_ipynb\n",
    "import definitions as ds\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.metrics import plot_confusion_matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reading Level 2 xlsx file, and adding the opposition statistics for a deeper analysis. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Charging Datasets for analysis\n",
    "#Level 1 (Fixing match Id for accounting both teams in the match.)\n",
    "Level2=pd.read_excel('Level2.xlsx')\n",
    "\n",
    "#Showing results after connecting opposition stats to compare with performance.\n",
    "Level2_df=ds.Opposites_To(Level2)\n",
    "\n",
    "#for just now, keeping out team Id indicator\n",
    "notsearchfor=['teamId_x','teamId_y','matchId_x','matchId_y','Others','shot','Goalkeeper','Save attempt','Shot']\n",
    "              #,'Save attempt','Goalkeeper','Others','Shot']\n",
    "Level2_df=Level2_df.drop(Level2_df.columns[Level2_df.columns.str.contains('')==Level2_df.columns.str.contains('|'.join(notsearchfor))],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Model Try-Out**: Trying out different model that could work well with classifying 1- Victory 0 - Draw/Defeat. Selection of 6 models, using ```GridSearchCV``` evaluating different parameters for each model depending on the matchweeks and stats from matches in the train set. \n",
    "\n",
    "In this case evaluated test window is moving, if \"Final Matchweek\" analyzed is f.ex: 30, test set will consist of the 3 following matchweek matches. Matches from MW: 31, 32, 33.\n",
    "<br>\n",
    "**Transformation_ByOrder function:** with the following parameters... \n",
    ">-**X**: Dataset of events to be transformed.\n",
    "<br>\n",
    ">-**Origin**: Index of the start of the analyzed set of events. \n",
    "<br>\n",
    ">-**Border**: Index of the end of the analyzed set of events. \n",
    "<br>\n",
    ">-**Test**: Index length of the test set of events. \n",
    "\n",
    "*For more info visit this link:*\n",
    "<br>\n",
    "**Modelling function**: with the following parameters...\n",
    ">**x**: feature train set to be modeled.\n",
    "<br>\n",
    ">**y**: target train set to be modeled.\n",
    "<br>\n",
    ">**xt**: feature test set to be modeled.\n",
    "<br>\n",
    ">**yt**: target test set to be modeled.\n",
    "<br>\n",
    "\n",
    "*For more info visit this link:\n",
    "<br>\n",
    "https://github.com/fabriziolufe/GRI-Research---Passing-Networks/blob/main/definitions.ipynb*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start Matchweek:  12.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  30.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  24.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  0.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  27.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  21.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  28.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  25.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  35.0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Training F1 Score</th>\n",
       "      <th>Training Accuracy</th>\n",
       "      <th>Testing F1 Score</th>\n",
       "      <th>Testing Accuracy</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Random Forest</th>\n",
       "      <td>0.770087</td>\n",
       "      <td>0.781404</td>\n",
       "      <td>0.699477</td>\n",
       "      <td>0.725439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SGD Classifier</th>\n",
       "      <td>0.778743</td>\n",
       "      <td>0.780377</td>\n",
       "      <td>0.712492</td>\n",
       "      <td>0.714912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM Classifier</th>\n",
       "      <td>0.799662</td>\n",
       "      <td>0.803316</td>\n",
       "      <td>0.704086</td>\n",
       "      <td>0.739474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Softmax Regression</th>\n",
       "      <td>0.803456</td>\n",
       "      <td>0.806254</td>\n",
       "      <td>0.732640</td>\n",
       "      <td>0.742105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Voting Classifier</th>\n",
       "      <td>0.784358</td>\n",
       "      <td>0.797605</td>\n",
       "      <td>0.685849</td>\n",
       "      <td>0.723684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XG Boost</th>\n",
       "      <td>0.830764</td>\n",
       "      <td>0.833219</td>\n",
       "      <td>0.726580</td>\n",
       "      <td>0.736842</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Training F1 Score  Training Accuracy  Testing F1 Score  \\\n",
       "Model                                                                        \n",
       "Random Forest                0.770087           0.781404          0.699477   \n",
       "SGD Classifier               0.778743           0.780377          0.712492   \n",
       "SVM Classifier               0.799662           0.803316          0.704086   \n",
       "Softmax Regression           0.803456           0.806254          0.732640   \n",
       "Voting Classifier            0.784358           0.797605          0.685849   \n",
       "XG Boost                     0.830764           0.833219          0.726580   \n",
       "\n",
       "                    Testing Accuracy  \n",
       "Model                                 \n",
       "Random Forest               0.725439  \n",
       "SGD Classifier              0.714912  \n",
       "SVM Classifier              0.739474  \n",
       "Softmax Regression          0.742105  \n",
       "Voting Classifier           0.723684  \n",
       "XG Boost                    0.736842  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#using Modelling function. Trying out grid search over models such as Softmax Regression, XGB, Random Forest, SGD and SVC. \n",
    "#For more details look over definitions.ipynb file.\n",
    "\n",
    "model_df = []\n",
    "i=0\n",
    "for i in range (1,20):\n",
    "    #Initializing train and test sets and showing first results. \n",
    "    x_train1, y_train1, x_test1, y_test1=ds.Transformation_ByOrder(Level2_df,ds.Mweeks_Y,ds.Mweeks_Z,[60])\n",
    "    Data=ds.Modelling(x_train1, y_train1, x_test1, y_test1)\n",
    "    #appending obtained data into an array\n",
    "    model_df.append(Data['Results'])\n",
    "    i=i+1\n",
    "    \n",
    "\n",
    "#Creating the dataframe.\n",
    "model_df = pd.concat(model_df).reset_index().drop('index',axis=1)  \n",
    "\n",
    "\n",
    "#showing results.\n",
    "model_df.groupby(['Model']).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Proceeding to evaluate two different models ( with best accuracy for the data...). First run with ```XGBClassifier()``` and Second run with ```LogisticRegression()```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**FIRST RUN (XGBOOST)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#best classifier\n",
    "# Training and testing sets (random)\n",
    "x_train, y_train, x_test, y_test=ds.Transformation_ByOrder(Level2_df,[0],[380],[60])\n",
    "\n",
    "##parameters for GridSearch\n",
    "parameters = {'learning_rate': [0.1, 0.15, 0.2, 0.25, 0.3, 0.35, 0.4],'n_estimators': [40,50],'max_depth': [2, 3, 4,None],\n",
    "              'colsample_bytree': [0.2,0.3,0.5,0.6,0.8],'alpha': [1e-5] }\n",
    "##creating a scorer for the model\n",
    "f1_scorer=ds.make_scorer(ds.f1_score, pos_label=1, average='weighted')\n",
    "\n",
    "#Grid Search\n",
    "XGB_fit=ds.xgb.XGBClassifier(eval_metric='mlogloss')\n",
    "grid_obj = ds.GridSearchCV(XGB_fit, scoring=f1_scorer, param_grid=parameters,cv=5).fit(x_train,y_train, eval_metric='mlogloss')\n",
    "\n",
    "##Obtaining best Estimator\n",
    "xgb_clf=grid_obj.best_estimator_\n",
    "#feature importance\n",
    "fit_coef=xgb_clf.feature_importances_\n",
    "print('Model Classifier: ', xgb_clf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use of the `Iteration_by_Matchweeks` function. \n",
    "<br>\n",
    "**Iteration_by_Matchweeks function:** Accounts for the following parameters..\n",
    "<br>\n",
    ">**Level_Dataset**: Dataset of grouped events at the selected level. 1,2 or 3.\n",
    "<br>\n",
    ">**Classifier**: Selected classifier to model.\n",
    "<br>\n",
    ">**Importance**: feature importance\n",
    "<br>\n",
    ">**Iterations**: 'int', # of iterations\n",
    "\n",
    "*For more info visit the link:\n",
    "<br>\n",
    "https://github.com/fabriziolufe/GRI-Research---Passing-Networks/blob/main/definitions.ipynb*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start Matchweek:  0.0 Final Matchweek:  19.0\n",
      "Model Classifier:  XGBClassifier(alpha=1e-05, base_score=0.5, booster='gbtree',\n",
      "              colsample_bylevel=1, colsample_bynode=1, colsample_bytree=0.2,\n",
      "              eval_metric='mlogloss', gamma=0, gpu_id=-1,\n",
      "              importance_type='gain', interaction_constraints='',\n",
      "              learning_rate=0.15, max_delta_step=0, max_depth=4,\n",
      "              min_child_weight=1, missing=nan, monotone_constraints='()',\n",
      "              n_estimators=50, n_jobs=12, num_parallel_tree=1, random_state=0,\n",
      "              reg_alpha=9.99999975e-06, reg_lambda=1, scale_pos_weight=1,\n",
      "              subsample=1, tree_method='exact', validate_parameters=1,\n",
      "              verbosity=None)\n",
      "Start Matchweek:  19.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  6.0\n",
      "Start Matchweek:  24.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  21.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  15.0 Final Matchweek:  20.0\n",
      "Start Matchweek:  29.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  24.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  21.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  21.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  20.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  0.0 Final Matchweek:  10.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  7.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  14.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  26.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  22.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  18.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  14.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  15.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  15.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  13.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  20.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  24.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  12.0\n",
      "Start Matchweek:  25.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  21.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  14.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  30.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  25.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  17.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  24.0\n",
      "Start Matchweek:  28.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  12.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  17.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  7.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  16.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  24.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  7.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  9.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  10.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  16.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  17.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  16.0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>accuracy in test</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top n_features</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.693833</td>\n",
       "      <td>0.673037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.693333</td>\n",
       "      <td>0.675101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.689667</td>\n",
       "      <td>0.670279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.689500</td>\n",
       "      <td>0.672614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.689000</td>\n",
       "      <td>0.668721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.688333</td>\n",
       "      <td>0.670169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.687500</td>\n",
       "      <td>0.669801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.687000</td>\n",
       "      <td>0.667753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.686000</td>\n",
       "      <td>0.667581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.686000</td>\n",
       "      <td>0.667061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.685833</td>\n",
       "      <td>0.668299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.683333</td>\n",
       "      <td>0.664971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>0.683333</td>\n",
       "      <td>0.666816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.683167</td>\n",
       "      <td>0.665005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.682500</td>\n",
       "      <td>0.663485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.682333</td>\n",
       "      <td>0.664949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.680667</td>\n",
       "      <td>0.662094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.680500</td>\n",
       "      <td>0.663127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.679833</td>\n",
       "      <td>0.661178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.678000</td>\n",
       "      <td>0.662724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.677833</td>\n",
       "      <td>0.660589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.676667</td>\n",
       "      <td>0.659788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.675833</td>\n",
       "      <td>0.650956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0.675167</td>\n",
       "      <td>0.657681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.674833</td>\n",
       "      <td>0.654620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>0.671667</td>\n",
       "      <td>0.653252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.671667</td>\n",
       "      <td>0.651218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>0.669500</td>\n",
       "      <td>0.652634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>0.669500</td>\n",
       "      <td>0.650540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>0.669333</td>\n",
       "      <td>0.649100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>0.669333</td>\n",
       "      <td>0.652619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>0.668500</td>\n",
       "      <td>0.652204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>0.668000</td>\n",
       "      <td>0.647428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>0.667833</td>\n",
       "      <td>0.646926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0.667333</td>\n",
       "      <td>0.651893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>0.666500</td>\n",
       "      <td>0.649759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>0.665500</td>\n",
       "      <td>0.650754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>0.665333</td>\n",
       "      <td>0.646989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>0.665000</td>\n",
       "      <td>0.648529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>0.664667</td>\n",
       "      <td>0.647427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.664167</td>\n",
       "      <td>0.647161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.663333</td>\n",
       "      <td>0.644007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>0.662000</td>\n",
       "      <td>0.643590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>0.661500</td>\n",
       "      <td>0.644372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>0.661167</td>\n",
       "      <td>0.643350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.659667</td>\n",
       "      <td>0.637169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.658667</td>\n",
       "      <td>0.639014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>0.657667</td>\n",
       "      <td>0.638997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.656833</td>\n",
       "      <td>0.637858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>0.653167</td>\n",
       "      <td>0.636146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.647833</td>\n",
       "      <td>0.627534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.633833</td>\n",
       "      <td>0.617509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.631833</td>\n",
       "      <td>0.613428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.626167</td>\n",
       "      <td>0.594645</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.616500</td>\n",
       "      <td>0.595605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.604000</td>\n",
       "      <td>0.550910</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                accuracy in test  F1 Score\n",
       "top n_features                            \n",
       "33                      0.693833  0.673037\n",
       "34                      0.693333  0.675101\n",
       "17                      0.689667  0.670279\n",
       "22                      0.689500  0.672614\n",
       "24                      0.689000  0.668721\n",
       "30                      0.688333  0.670169\n",
       "23                      0.687500  0.669801\n",
       "32                      0.687000  0.667753\n",
       "16                      0.686000  0.667581\n",
       "31                      0.686000  0.667061\n",
       "15                      0.685833  0.668299\n",
       "14                      0.683333  0.664971\n",
       "35                      0.683333  0.666816\n",
       "13                      0.683167  0.665005\n",
       "21                      0.682500  0.663485\n",
       "27                      0.682333  0.664949\n",
       "12                      0.680667  0.662094\n",
       "26                      0.680500  0.663127\n",
       "18                      0.679833  0.661178\n",
       "10                      0.678000  0.662724\n",
       "20                      0.677833  0.660589\n",
       "25                      0.676667  0.659788\n",
       "19                      0.675833  0.650956\n",
       "55                      0.675167  0.657681\n",
       "29                      0.674833  0.654620\n",
       "47                      0.671667  0.653252\n",
       "28                      0.671667  0.651218\n",
       "56                      0.669500  0.652634\n",
       "51                      0.669500  0.650540\n",
       "54                      0.669333  0.649100\n",
       "40                      0.669333  0.652619\n",
       "44                      0.668500  0.652204\n",
       "53                      0.668000  0.647428\n",
       "52                      0.667833  0.646926\n",
       "41                      0.667333  0.651893\n",
       "36                      0.666500  0.649759\n",
       "42                      0.665500  0.650754\n",
       "48                      0.665333  0.646989\n",
       "43                      0.665000  0.648529\n",
       "45                      0.664667  0.647427\n",
       "11                      0.664167  0.647161\n",
       "50                      0.663333  0.644007\n",
       "39                      0.662000  0.643590\n",
       "46                      0.661500  0.644372\n",
       "49                      0.661167  0.643350\n",
       "9                       0.659667  0.637169\n",
       "8                       0.658667  0.639014\n",
       "37                      0.657667  0.638997\n",
       "7                       0.656833  0.637858\n",
       "38                      0.653167  0.636146\n",
       "6                       0.647833  0.627534\n",
       "4                       0.633833  0.617509\n",
       "5                       0.631833  0.613428\n",
       "2                       0.626167  0.594645\n",
       "3                       0.616500  0.595605\n",
       "1                       0.604000  0.550910"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#XGB Results\n",
    "XGBdf=ds.Iteration_by_Matchweeks(Level2_df, xgb_clf,fit_coef,100)\n",
    "XGBdf.groupby(['top n_features'])['accuracy in test', 'F1 Score'].mean().sort_values(by='accuracy in test',ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After getting results we want to evaluate which were the most important features that accounted for a better model accuracy. By using ```matplotlib.pyplot``` we can generate a Horizonal Bar graph with results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feat_importances =pd.Series(fit_coef, index=x_train.columns)\n",
    "data=np.power(feat_importances,2)\n",
    "dataSq=data.nlargest(10)\n",
    "Xtop=dataSq.index.tolist()\n",
    "\n",
    "#visualization of feature importance\n",
    "feat_importances[Xtop].plot(kind='barh')\n",
    "ds.plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**SECOND RUN (Logistic Regression)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start Matchweek:  0.0 Final Matchweek:  19.0\n",
      "Model Classifier:  LogisticRegression(C=10, penalty='l1', solver='liblinear')\n",
      "Start Matchweek:  8.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  13.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  17.0\n",
      "Start Matchweek:  22.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  0.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  12.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  8.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  20.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  27.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  7.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  11.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  17.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  11.0\n",
      "Start Matchweek:  21.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  21.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  20.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  13.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  7.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  26.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  18.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  18.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  17.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  21.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  9.0\n",
      "Start Matchweek:  13.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  29.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  15.0 Final Matchweek:  20.0\n",
      "Start Matchweek:  13.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  16.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  13.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  15.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  14.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  22.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  0.0 Final Matchweek:  15.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  13.0 Final Matchweek:  33.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  11.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  31.0\n",
      "Start Matchweek:  10.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  12.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  24.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  27.0 Final Matchweek:  32.0\n",
      "Start Matchweek:  4.0 Final Matchweek:  19.0\n",
      "Start Matchweek:  5.0 Final Matchweek:  30.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  18.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  27.0\n",
      "Start Matchweek:  23.0 Final Matchweek:  28.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  6.0 Final Matchweek:  21.0\n",
      "Start Matchweek:  30.0 Final Matchweek:  35.0\n",
      "Start Matchweek:  7.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  16.0\n",
      "Start Matchweek:  8.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  17.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  2.0 Final Matchweek:  22.0\n",
      "Start Matchweek:  29.0 Final Matchweek:  34.0\n",
      "Start Matchweek:  3.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  18.0 Final Matchweek:  23.0\n",
      "Start Matchweek:  1.0 Final Matchweek:  16.0\n",
      "Start Matchweek:  24.0 Final Matchweek:  29.0\n",
      "Start Matchweek:  11.0 Final Matchweek:  21.0\n",
      "Start Matchweek:  15.0 Final Matchweek:  25.0\n",
      "Start Matchweek:  16.0 Final Matchweek:  26.0\n",
      "Start Matchweek:  0.0 Final Matchweek:  20.0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>accuracy in test</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top n_features</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.725667</td>\n",
       "      <td>0.723467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.722500</td>\n",
       "      <td>0.720400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.714333</td>\n",
       "      <td>0.712927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.713500</td>\n",
       "      <td>0.711448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.711500</td>\n",
       "      <td>0.710245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.711167</td>\n",
       "      <td>0.685893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.709833</td>\n",
       "      <td>0.707716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.709500</td>\n",
       "      <td>0.685692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>0.707833</td>\n",
       "      <td>0.705061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>0.707500</td>\n",
       "      <td>0.705444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.706833</td>\n",
       "      <td>0.686935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.706833</td>\n",
       "      <td>0.704443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.706167</td>\n",
       "      <td>0.704122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.704167</td>\n",
       "      <td>0.702194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.703667</td>\n",
       "      <td>0.701679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.702667</td>\n",
       "      <td>0.695257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.701667</td>\n",
       "      <td>0.699907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>0.701000</td>\n",
       "      <td>0.698529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.700667</td>\n",
       "      <td>0.691076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>0.700333</td>\n",
       "      <td>0.698100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>0.700167</td>\n",
       "      <td>0.697668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.698777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>0.699500</td>\n",
       "      <td>0.697875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.699333</td>\n",
       "      <td>0.693574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.698667</td>\n",
       "      <td>0.694311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.697500</td>\n",
       "      <td>0.694491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0.697500</td>\n",
       "      <td>0.696424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.695833</td>\n",
       "      <td>0.678930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.694833</td>\n",
       "      <td>0.677011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.694833</td>\n",
       "      <td>0.680893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.694667</td>\n",
       "      <td>0.685725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>0.694333</td>\n",
       "      <td>0.693148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.693667</td>\n",
       "      <td>0.685959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.693333</td>\n",
       "      <td>0.685640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.692667</td>\n",
       "      <td>0.689482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>0.692333</td>\n",
       "      <td>0.690818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.690667</td>\n",
       "      <td>0.688633</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.690667</td>\n",
       "      <td>0.688896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.690333</td>\n",
       "      <td>0.686771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>0.689833</td>\n",
       "      <td>0.689671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.688333</td>\n",
       "      <td>0.677790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>0.686833</td>\n",
       "      <td>0.686683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>0.685000</td>\n",
       "      <td>0.684448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.685000</td>\n",
       "      <td>0.681442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>0.684333</td>\n",
       "      <td>0.683396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>0.684167</td>\n",
       "      <td>0.683250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>0.684167</td>\n",
       "      <td>0.683250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>0.683667</td>\n",
       "      <td>0.683416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0.683333</td>\n",
       "      <td>0.681808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>0.678167</td>\n",
       "      <td>0.675686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>0.677000</td>\n",
       "      <td>0.676381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.676833</td>\n",
       "      <td>0.676062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>0.676667</td>\n",
       "      <td>0.675914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.644500</td>\n",
       "      <td>0.518684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.640833</td>\n",
       "      <td>0.516697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.637833</td>\n",
       "      <td>0.516927</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                accuracy in test  F1 Score\n",
       "top n_features                            \n",
       "24                      0.725667  0.723467\n",
       "23                      0.722500  0.720400\n",
       "25                      0.714333  0.712927\n",
       "33                      0.713500  0.711448\n",
       "26                      0.711500  0.710245\n",
       "4                       0.711167  0.685893\n",
       "27                      0.709833  0.707716\n",
       "5                       0.709500  0.685692\n",
       "36                      0.707833  0.705061\n",
       "35                      0.707500  0.705444\n",
       "6                       0.706833  0.686935\n",
       "22                      0.706833  0.704443\n",
       "34                      0.706167  0.704122\n",
       "21                      0.704167  0.702194\n",
       "28                      0.703667  0.701679\n",
       "15                      0.702667  0.695257\n",
       "29                      0.701667  0.699907\n",
       "38                      0.701000  0.698529\n",
       "10                      0.700667  0.691076\n",
       "37                      0.700333  0.698100\n",
       "39                      0.700167  0.697668\n",
       "42                      0.700000  0.698777\n",
       "40                      0.699500  0.697875\n",
       "16                      0.699333  0.693574\n",
       "17                      0.698667  0.694311\n",
       "30                      0.697500  0.694491\n",
       "41                      0.697500  0.696424\n",
       "8                       0.695833  0.678930\n",
       "7                       0.694833  0.677011\n",
       "9                       0.694833  0.680893\n",
       "12                      0.694667  0.685725\n",
       "44                      0.694333  0.693148\n",
       "13                      0.693667  0.685959\n",
       "14                      0.693333  0.685640\n",
       "18                      0.692667  0.689482\n",
       "43                      0.692333  0.690818\n",
       "31                      0.690667  0.688633\n",
       "32                      0.690667  0.688896\n",
       "19                      0.690333  0.686771\n",
       "45                      0.689833  0.689671\n",
       "11                      0.688333  0.677790\n",
       "46                      0.686833  0.686683\n",
       "51                      0.685000  0.684448\n",
       "20                      0.685000  0.681442\n",
       "54                      0.684333  0.683396\n",
       "52                      0.684167  0.683250\n",
       "53                      0.684167  0.683250\n",
       "47                      0.683667  0.683416\n",
       "55                      0.683333  0.681808\n",
       "56                      0.678167  0.675686\n",
       "48                      0.677000  0.676381\n",
       "50                      0.676833  0.676062\n",
       "49                      0.676667  0.675914\n",
       "1                       0.644500  0.518684\n",
       "2                       0.640833  0.516697\n",
       "3                       0.637833  0.516927"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#best classifier\n",
    "## Training and testing sets\n",
    "#19 first games\n",
    "x_train2, y_train2, x_test2, y_test2=ds.Transformation_ByOrder(Level2_df,[0],[380],[100])\n",
    "\n",
    "#Logistic Regression\n",
    "##Parameters\n",
    "modelLR= ds.LogisticRegression()\n",
    "m_class=['auto','ovr','multinomial']\n",
    "solvers = ['newton-cg', 'lbfgs', 'liblinear']\n",
    "penalty = ['l1','l2']\n",
    "c_values = [100, 10, 1.0, 0.1, 0.01]\n",
    "# define grid search\n",
    "gridLR= dict(multi_class=m_class,solver=solvers,penalty=penalty,C=c_values)\n",
    "cvLR= ds.RepeatedStratifiedKFold(n_splits=10, n_repeats=3, random_state=1)\n",
    "grid_searchLR= ds.GridSearchCV(estimator=modelLR, param_grid=gridLR, n_jobs=-1, cv=cvLR, scoring='accuracy',error_score=0)\n",
    "softmax_grid = grid_searchLR.fit(x_train2, y_train2)\n",
    "\n",
    "#best estimator for the grid search\n",
    "softmax_reg= softmax_grid.best_estimator_\n",
    "#feature importance\n",
    "coefficients=softmax_reg.coef_[0]\n",
    "print('Model Classifier: ', softmax_reg)\n",
    "\n",
    "LRdf=ds.Iteration_by_Matchweeks(Level2_df, softmax_reg,coefficients,100)\n",
    "LRdf.groupby(['top n_features'])['accuracy in test', 'F1 Score'].mean().sort_values(by='accuracy in test',ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After getting results we want to evaluate which were the most important features that accounted for a better model accuracy. By using ```matplotlib.pyplot``` we can generate a Horizonal Bar graph with results. ```sklearn.metrics``` creates a confusion matrix which guide us into knowing how the classifier works in every specific case."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Feature importances and the feature names assigned. \n",
    "feat_importances =pd.Series(coefficients, index=x_train2.columns)\n",
    "\n",
    "#in order to get both negative and possitives as equal, we transform to the power of 2.\n",
    "data=np.power(feat_importances,2)\n",
    "dataSq=data.nlargest(10)\n",
    "Xtop=dataSq.index.tolist()\n",
    "\n",
    "#visualization of feature importance\n",
    "feat_importances[Xtop].plot(kind='barh')\n",
    "ds.plt.show()\n",
    "\n",
    "x_test_new=x_train2[Xtop]\n",
    "model_clf=softmax_reg\n",
    "#confusion matrix\n",
    "plot_confusion_matrix(model_clf, x_test_new, y_test2)  \n",
    "ds.plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case by having more depth inside the features, the classifier managed to raise it's accuracy of assigning when a team won up to 46.9%. The other case ( when a team lose - (0)) get 79.4% accuracy, with the most important feature being the amount of crosses made during the game. A team crossing too much can be a disadvantage for open play strategy and goal building. As we search through literature we found that some studies have been made in the subject: Crossing in Soccer has a Strong Negative Impact on Scoring: Evidence from the English Premier League the German Bundesliga and the World Cup 2014"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![emery.pg](emery.jpg \"Title\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Link to Original Dataset documentation**:\n",
    "    https://figshare.com/collections/Soccer_match_event_dataset/4415000/5\n",
    "\n",
    "**Published article from where Dataset was retrieved**:\n",
    "Pappalardo, L., Cintia, P., Rossi, A., Massucco, E., Ferragina, P., Pedreschi, D., &amp; Giannotti, F. (2019). A public data  set of spatio-temporal match events in soccer competitions. Scientific Data, 6(1). https://doi.org/10.1038/s41597-019-0247-7 "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}