GridSearchCV.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "from sklearn import linear_model\n",
    "from sklearn.datasets import make_classification\n",
    "from collections import Counter \n",
    "from sklearn.ensemble import GradientBoostingClassifier, RandomForestClassifier\n",
    "from sklearn.preprocessing import StandardScaler, LabelBinarizer\n",
    "from sklearn.metrics import (roc_curve, precision_recall_curve, auc, make_scorer, classification_report, \n",
    "                             recall_score, accuracy_score, precision_score, confusion_matrix)\n",
    "from sklearn.model_selection import RandomizedSearchCV,train_test_split, GridSearchCV, cross_val_score, KFold\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.pipeline import Pipeline\n",
    "from mlxtend.plotting import plot_decision_regions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dados em cada uma das classes [(0, 68), (1, 4932)]\n"
     ]
    }
   ],
   "source": [
    "X,y =make_classification(n_samples=5000, n_features=6, n_informative=2, n_classes=2, n_clusters_per_class=1,\n",
    "                        n_redundant=0,n_repeated=0,weights=[0.01,0.99],class_sep=0.8,random_state=0)\n",
    "print(\"Dados em cada uma das classes\",sorted(Counter(y).items()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3750, 6) (1250, 6)\n",
      "(3750,) (1250,)\n"
     ]
    }
   ],
   "source": [
    "x_train,x_test=train_test_split(X)\n",
    "print(x_train.shape,x_test.shape)\n",
    "y_train,y_test=train_test_split(y)\n",
    "print(y_train.shape,y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Muitas vezes temos um algoritmo para solucionar um problema de classificação ou regressão mas o mesmo possui diversos parâmetros que, se combinados de diferentes formas, podem produzir resultados distintos. Nesse caso, para testar esses diferentes parâmetros temos algumas alternativas. A priemeira seria, manualmente, tentar diferentes valores para esses parâmetros, pegar os resultados finais e definir qual seria o melhor modelo. Essa solução parece ser pouco eficiente na medida que exige muito esforço e um enorme risco de calibração, em especial para parâmetros contínuos. A segunda solução seria fazer esse ajuste fino usando o GridSearchCV. Nesse caso, testamos diferentes valores para os hiperparâmetros e obtemos o classificador que apresenta o maior score para recall. Por fim, podemos ajustar nosso hiperparâmetro considerando como métrica a curva precision-recall ou roc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GridSearchCV"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vamos ver como aplicar o GridSearchCV a um simples problema para o estimador svm. Veja abaixo que esse estimador possui diversos parâmetros, em especial queremos encontrar qual seria o melhor parâmetro para C, degree e a função kernel. Veja que definimos o estimador e depois o intervalo de oscilação dos parâmetros. A seguir, definimos o GridSearchCV."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto_deprecated',\n",
       "  kernel='rbf', max_iter=-1, probability=False, random_state=None,\n",
       "  shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SVC()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVC(C=1, cache_size=200, class_weight=None, coef0=0.0,\n",
      "  decision_function_shape='ovr', degree=1, gamma='auto_deprecated',\n",
      "  kernel='linear', max_iter=-1, probability=True, random_state=None,\n",
      "  shrinking=True, tol=0.001, verbose=False)\n",
      "0.988\n",
      "{'C': 1, 'degree': 1, 'kernel': 'linear'}\n",
      "{'C': 1, 'degree': 1, 'kernel': 'linear'}\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "clf1 = SVC(probability=True)   #classificador\n",
    "params1 = {'C': [1, 10, 100],\"degree\":[1,2,3],\"kernel\":('linear',\"rbf\")}   #parametros\n",
    "gs1=GridSearchCV(clf1,params1,cv=5,n_jobs=-1,error_score=0)\n",
    "gs1.fit(x_train,y_train)\n",
    "y_pred=gs1.predict(x_test)\n",
    "y_prob=gs1.predict_proba(x_test)\n",
    "print(gs1.best_estimator_)\n",
    "print(gs1.best_score_)\n",
    "print(gs1.best_params_)\n",
    "print(gs1.cv_results_['params'][gs1.best_index_])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "================================="
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Agora podemos testar outro estimador, o RandomForestClassifier, que possui um conjunto diferente de hiperparâmetros. Primeiro vemos como é o estimador. Vamos testar diferentes combinações com \"min_samples_split\", \"n_estimators\", \"max_depth\" e \" max_features\". Mas, além disso, vamos ver como seria com diferentes métricas para o score."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators='warn', n_jobs=None,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RandomForestClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
      "            max_depth=15, max_features=2, max_leaf_nodes=None,\n",
      "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "            min_samples_leaf=1, min_samples_split=12,\n",
      "            min_weight_fraction_leaf=0.0, n_estimators=300, n_jobs=None,\n",
      "            oob_score=False, random_state=None, verbose=0,\n",
      "            warm_start=False)\n",
      "0.5710301394511921\n",
      "{'max_depth': 15, 'max_features': 2, 'min_samples_split': 12, 'n_estimators': 300}\n"
     ]
    }
   ],
   "source": [
    "clf2 = RandomForestClassifier()\n",
    "params2 = {'min_samples_split':range(2, 32, 10),'n_estimators':[100, 300],\n",
    "           'max_depth':[3, 5, 15, 25],'max_features':[2,4,6]}\n",
    "scor={\"AUC\":'roc_auc',\"Accuracy\":make_scorer(accuracy_score)}\n",
    "gs2=GridSearchCV(clf2,param_grid=params2,scoring=scor,cv=5,refit=\"AUC\",\n",
    "                 return_train_score=True,n_jobs=-1,error_score=0)\n",
    "gs2.fit(x_train,y_train)\n",
    "print(gs2.best_estimator_)\n",
    "print(gs2.best_score_)\n",
    "print(gs2.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "resultados=gs2.cv_results_\n",
    "plt.figure(figsize=(10,10))\n",
    "\n",
    "plt.xlabel(\"min_samples_split\")\n",
    "plt.ylabel(\"Score\")\n",
    "\n",
    "ax = plt.gca()\n",
    "ax.set_xlim(0, 30)\n",
    "ax.set_ylim(0.90, 1)\n",
    "X_axis = np.array(resultados['param_min_samples_split'].data, dtype=float)\n",
    "for scorer, color in zip(sorted(scor), ['g', 'k']):\n",
    "    for sample, style in (('train', '--'), ('test', '-')):\n",
    "        sample_score_mean = resultados['mean_%s_%s' % (sample, scorer)]\n",
    "        sample_score_std = resultados['std_%s_%s' % (sample, scorer)]\n",
    "        ax.fill_between(X_axis, sample_score_mean - sample_score_std,\n",
    "                        sample_score_mean + sample_score_std,\n",
    "                        alpha=0.1 if sample == 'test' else 0, color=color)\n",
    "        ax.plot(X_axis, sample_score_mean, style, color=color,\n",
    "                alpha=1 if sample == 'test' else 0.9,\n",
    "                label=\"%s (%s)\" % (scorer, sample))\n",
    "\n",
    "    best_index = np.nonzero(resultados['rank_test_%s' % scorer] == 1)[0][0]\n",
    "    best_score = resultados['mean_test_%s' % scorer][best_index]\n",
    "\n",
    "    # Plot a dotted vertical line at the best score for that scorer marked by x\n",
    "    ax.plot([X_axis[best_index], ] * 2, [0, best_score],\n",
    "            linestyle='-.', color=color, marker='x', markeredgewidth=3, ms=8)\n",
    "\n",
    "    # Annotate the best score for that scorer\n",
    "    ax.annotate(\"%0.2f\" % best_score,\n",
    "                (X_axis[best_index], best_score + 0.005))\n",
    "\n",
    "plt.legend(loc=\"best\")\n",
    "plt.grid('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "==========================="
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Agora vamos ver como podemos otimizar um estimador (classifier) usando o recurso do cross-validation. Para tanto vamos especificar o estimador SVC para diferentes hiperparâmetros e scores."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'C': 1, 'degree': 1, 'kernel': 'linear'}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.00      0.00      0.00        23\n",
      "           1       0.98      1.00      0.99      1227\n",
      "\n",
      "   micro avg       0.98      0.98      0.98      1250\n",
      "   macro avg       0.49      0.50      0.50      1250\n",
      "weighted avg       0.96      0.98      0.97      1250\n",
      "\n",
      "[[   0   23]\n",
      " [   0 1227]]\n",
      "{'C': 1, 'degree': 1, 'kernel': 'linear'}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.00      0.00      0.00        23\n",
      "           1       0.98      1.00      0.99      1227\n",
      "\n",
      "   micro avg       0.98      0.98      0.98      1250\n",
      "   macro avg       0.49      0.50      0.50      1250\n",
      "weighted avg       0.96      0.98      0.97      1250\n",
      "\n",
      "[[   0   23]\n",
      " [   0 1227]]\n"
     ]
    }
   ],
   "source": [
    "clf3 = SVC(probability=True)   #classificador\n",
    "params1 = {'C': [1, 10, 100],\"degree\":[1,2,3],\"kernel\":('linear',\"rbf\")}   #parametros\n",
    "scores=[\"precision\",\"recall\"]\n",
    "for j in scores:\n",
    "    classif=GridSearchCV(clf3,params1,cv=5,scoring=j)\n",
    "    classif.fit(x_train,y_train)\n",
    "    print(classif.best_params_)\n",
    "    y_pred3=classif.predict(x_test)\n",
    "    print(classification_report(y_test,y_pred))\n",
    "    print(confusion_matrix(y_test,y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Search\n",
    "Esse método implementa uma busca aleatória dos parâmetros ótimos a partir de uma seleção de possíveis valores em uma distibuição."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'shuffle': True, 'n_iter': 69, 'alpha': 0.1} 0.9348\n"
     ]
    }
   ],
   "source": [
    "random = RandomizedSearchCV(estimator=clf, param_distributions=params, cv = 5, n_jobs=-1)\n",
    "RS=random.fit(X,y)\n",
    "print(RS.best_params_,RS.best_score_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PIPELINE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "O recurso \"pipeline\" é uma classe que possui uma ferramenta bem útil para agregar diferentes \"transformadores\" em um estimador dentro de um objeto e que possui dois importantes componentes: <br>\n",
    "1. Transformers: \n",
    "  Implementa tanto o fit() quanto o transform().\n",
    "2. Estimators:\n",
    "  Implementa o fit() e o predict()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Um pipeline simples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import f1_score,classification_report, confusion_matrix\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.svm import LinearSVC\n",
    "from mlxtend.plotting import plot_decision_regions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.14      0.25        64\n",
      "           1       0.81      0.72      0.76       262\n",
      "           2       0.98      1.00      0.99      4674\n",
      "\n",
      "   micro avg       0.97      0.97      0.97      5000\n",
      "   macro avg       0.93      0.62      0.67      5000\n",
      "weighted avg       0.97      0.97      0.97      5000\n",
      "\n",
      "[[   9   43   12]\n",
      " [   0  189   73]\n",
      " [   0    2 4672]]\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "pipeline = Pipeline([\n",
    "    ('scaler',StandardScaler()),\n",
    "    ('clf', LogisticRegression())])\n",
    "p1=pipeline.fit(X,y)\n",
    "y_pred=p1.predict(X)\n",
    "print(classification_report(y, y_pred))\n",
    "print(confusion_matrix(y,y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11d7fd9b0>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plot_decision_regions(X=X,y=y_pred,clf=p1,legend=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### pipeline com Keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.layers import Dense, Dropout\n",
    "from keras.models import Model, Sequential\n",
    "from keras.wrappers.scikit_learn import KerasClassifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Primeiro vamos construir o modelo que irá ser utilizado na estimativa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_model(optimizer='adagrad',kernel_initializer='glorot_uniform', dropout=0.2):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(64,activation='relu',kernel_initializer=kernel_initializer))\n",
    "    model.add(Dropout(dropout))\n",
    "    model.add(Dense(1,activation='sigmoid',kernel_initializer=kernel_initializer))\n",
    "    model.compile(loss='binary_crossentropy',optimizer=optimizer, metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Quando vamos usar o pipeline, lembre-se que temos duas componentes, o transformer, que no nosso caso seria mudar a escala dos dados \"StandardScaler()\", e o estimador, que para esse exemplo seria o \"KerasClassifier\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()    #transform\n",
    "keras1 = KerasClassifier(build_fn=create_model,verbose=0)   #fit e predict (nosso estimador)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid = {\n",
    "    'clf__optimizer':['adam','adagrad'],\n",
    "    'clf__epochs':[10,30],\n",
    "    'clf__dropout':[0.1,0.2],\n",
    "    'clf__kernel_initializer':['normal','uniform']\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "pipeline = Pipeline([\n",
    "    ('preprocess',scaler),\n",
    "    ('clf',keras1)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "grid = GridSearchCV(pipeline, cv=3, param_grid=param_grid,n_jobs=-1)\n",
    "p2=grid.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# summarize results\n",
    "print(\"Best: %f using %s\" % (p2.best_score_, p2.best_params_))\n",
    "means = p2.cv_results_['mean_test_score']\n",
    "stds = p2.cv_results_['std_test_score']\n",
    "params = p2.cv_results_['params']\n",
    "for mean, stdev, param in zip(means, stds, params):\n",
    "    print(\"%f (%f) with: %r\" % (mean, stdev, param))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred=p2.predict(X)\n",
    "plot_decision_regions(X=X,y=y,clf=svm1,legend=2)\n",
    "print(confusion_matrix(y,y_pred))"
   ]
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}