Hyper parameter Tuning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "7f16f7fc-e466-4f35-a570-487445dc8803",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import svm, datasets\n",
    "iris = datasets.load_iris()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e5c8e041-ae53-4eed-8759-a929cd8d5917",
   "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>sepal length (cm)</th>\n",
       "      <th>sepal width (cm)</th>\n",
       "      <th>petal length (cm)</th>\n",
       "      <th>petal width (cm)</th>\n",
       "      <th>flower</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>4.6</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>5.3</td>\n",
       "      <td>3.7</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>5.0</td>\n",
       "      <td>3.3</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>7.0</td>\n",
       "      <td>3.2</td>\n",
       "      <td>4.7</td>\n",
       "      <td>1.4</td>\n",
       "      <td>versicolor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>6.4</td>\n",
       "      <td>3.2</td>\n",
       "      <td>4.5</td>\n",
       "      <td>1.5</td>\n",
       "      <td>versicolor</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sepal length (cm)  sepal width (cm)  petal length (cm)  petal width (cm)  \\\n",
       "47                4.6               3.2                1.4               0.2   \n",
       "48                5.3               3.7                1.5               0.2   \n",
       "49                5.0               3.3                1.4               0.2   \n",
       "50                7.0               3.2                4.7               1.4   \n",
       "51                6.4               3.2                4.5               1.5   \n",
       "\n",
       "        flower  \n",
       "47      setosa  \n",
       "48      setosa  \n",
       "49      setosa  \n",
       "50  versicolor  \n",
       "51  versicolor  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df = pd.DataFrame(iris.data,columns=iris.feature_names)\n",
    "df['flower'] = iris.target\n",
    "df['flower'] = df['flower'].apply(lambda x: iris.target_names[x])\n",
    "df[47:52]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e15df0da-9f6e-4025-8bcb-b51cadc377a2",
   "metadata": {},
   "source": [
    "# Approach 1: Use train_test_split and manually tune parameters by trial and error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6462ba22-c6e0-43c9-bc98-5d8edecc1b3e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(iris.data, iris.target, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d182f176-bb41-4ade-8a27-80c3fc4b7661",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9777777777777777"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# by taking random parameter\n",
    "model = svm.SVC(kernel='rbf',C=30,gamma='auto')\n",
    "model.fit(X_train,y_train)\n",
    "model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd7fdc24-f335-4e80-ac4f-92162c3715c9",
   "metadata": {},
   "source": [
    "# Approach 2: Use K Fold Cross validation\r",
    "# \n",
    "Manually try suppling models with different parameters to cross_val_score function with 5 fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "7791043c-d957-4cb3-b9fb-982fb1d4ec77",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.        , 1.        , 0.9       , 0.96666667, 1.        ])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "cross_val_score(svm.SVC(kernel='linear',C=10,gamma='auto'),iris.data, iris.target, cv=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3929d6b8-8476-43a1-b28d-1d71e127459c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.96666667, 1.        , 0.96666667, 0.96666667, 1.        ])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(svm.SVC(kernel='rbf',C=10,gamma='auto'),iris.data, iris.target, cv=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "40499ed0-3200-4356-9f9e-c51cc9b0b3f6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.96666667, 1.        , 0.9       , 0.96666667, 1.        ])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(svm.SVC(kernel='rbf',C=20,gamma='auto'),iris.data, iris.target, cv=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8b87b4af-331e-4877-9d61-08e30a24fd49",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'rbf_1': 0.9800000000000001,\n",
       " 'rbf_10': 0.9800000000000001,\n",
       " 'rbf_20': 0.9666666666666668,\n",
       " 'linear_1': 0.9800000000000001,\n",
       " 'linear_10': 0.9733333333333334,\n",
       " 'linear_20': 0.9666666666666666}"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking which parameter will give me the best score\n",
    "import numpy as np\n",
    "kernels = ['rbf', 'linear']\n",
    "C = [1,10,20]\n",
    "avg_scores = {}\n",
    "for kval in kernels:\n",
    "    for cval in C:\n",
    "        cv_scores = cross_val_score(svm.SVC(kernel=kval,C=cval,gamma='auto'),iris.data, iris.target, cv=5)\n",
    "        avg_scores[kval + '_' + str(cval)] = np.average(cv_scores)\n",
    "\n",
    "avg_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a076e504-cbf0-4de0-817a-43306e20ceb0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'mean_fit_time': array([0.0016242 , 0.0014914 , 0.00117164, 0.00120392, 0.00111942,\n",
       "        0.00090656]),\n",
       " 'std_fit_time': array([0.00050429, 0.00044672, 0.00033951, 0.00075195, 0.0002371 ,\n",
       "        0.00019258]),\n",
       " 'mean_score_time': array([0.00153861, 0.00080185, 0.0009346 , 0.00102787, 0.00089083,\n",
       "        0.00100107]),\n",
       " 'std_score_time': array([4.73780865e-04, 4.00933473e-04, 5.35326328e-04, 3.53361572e-04,\n",
       "        1.89108026e-04, 1.15207252e-05]),\n",
       " 'param_C': masked_array(data=[1, 1, 10, 10, 20, 20],\n",
       "              mask=[False, False, False, False, False, False],\n",
       "        fill_value='?',\n",
       "             dtype=object),\n",
       " 'param_kernel': masked_array(data=['rbf', 'linear', 'rbf', 'linear', 'rbf', 'linear'],\n",
       "              mask=[False, False, False, False, False, False],\n",
       "        fill_value='?',\n",
       "             dtype=object),\n",
       " 'params': [{'C': 1, 'kernel': 'rbf'},\n",
       "  {'C': 1, 'kernel': 'linear'},\n",
       "  {'C': 10, 'kernel': 'rbf'},\n",
       "  {'C': 10, 'kernel': 'linear'},\n",
       "  {'C': 20, 'kernel': 'rbf'},\n",
       "  {'C': 20, 'kernel': 'linear'}],\n",
       " 'split0_test_score': array([0.96666667, 0.96666667, 0.96666667, 1.        , 0.96666667,\n",
       "        1.        ]),\n",
       " 'split1_test_score': array([1., 1., 1., 1., 1., 1.]),\n",
       " 'split2_test_score': array([0.96666667, 0.96666667, 0.96666667, 0.9       , 0.9       ,\n",
       "        0.9       ]),\n",
       " 'split3_test_score': array([0.96666667, 0.96666667, 0.96666667, 0.96666667, 0.96666667,\n",
       "        0.93333333]),\n",
       " 'split4_test_score': array([1., 1., 1., 1., 1., 1.]),\n",
       " 'mean_test_score': array([0.98      , 0.98      , 0.98      , 0.97333333, 0.96666667,\n",
       "        0.96666667]),\n",
       " 'std_test_score': array([0.01632993, 0.01632993, 0.01632993, 0.03887301, 0.03651484,\n",
       "        0.0421637 ]),\n",
       " 'rank_test_score': array([1, 1, 1, 4, 5, 6])}"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# doing the exact same thing with more effeciently\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "clf = GridSearchCV(svm.SVC(gamma='auto'), {\n",
    "    'C': [1,10,20],\n",
    "    'kernel': ['rbf','linear']\n",
    "}, cv=5, return_train_score=False)\n",
    "clf.fit(iris.data, iris.target)\n",
    "clf.cv_results_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d755ea8d-51ee-4b58-8bbf-a27bdff10143",
   "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>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_C</th>\n",
       "      <th>param_kernel</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>split3_test_score</th>\n",
       "      <th>split4_test_score</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.001624</td>\n",
       "      <td>0.000504</td>\n",
       "      <td>0.001539</td>\n",
       "      <td>0.000474</td>\n",
       "      <td>1</td>\n",
       "      <td>rbf</td>\n",
       "      <td>{'C': 1, 'kernel': 'rbf'}</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>0.016330</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.001491</td>\n",
       "      <td>0.000447</td>\n",
       "      <td>0.000802</td>\n",
       "      <td>0.000401</td>\n",
       "      <td>1</td>\n",
       "      <td>linear</td>\n",
       "      <td>{'C': 1, 'kernel': 'linear'}</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>0.016330</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.001172</td>\n",
       "      <td>0.000340</td>\n",
       "      <td>0.000935</td>\n",
       "      <td>0.000535</td>\n",
       "      <td>10</td>\n",
       "      <td>rbf</td>\n",
       "      <td>{'C': 10, 'kernel': 'rbf'}</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>0.016330</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.001204</td>\n",
       "      <td>0.000752</td>\n",
       "      <td>0.001028</td>\n",
       "      <td>0.000353</td>\n",
       "      <td>10</td>\n",
       "      <td>linear</td>\n",
       "      <td>{'C': 10, 'kernel': 'linear'}</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.973333</td>\n",
       "      <td>0.038873</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.001119</td>\n",
       "      <td>0.000237</td>\n",
       "      <td>0.000891</td>\n",
       "      <td>0.000189</td>\n",
       "      <td>20</td>\n",
       "      <td>rbf</td>\n",
       "      <td>{'C': 20, 'kernel': 'rbf'}</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.036515</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.000907</td>\n",
       "      <td>0.000193</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000012</td>\n",
       "      <td>20</td>\n",
       "      <td>linear</td>\n",
       "      <td>{'C': 20, 'kernel': 'linear'}</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.042164</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean_fit_time  std_fit_time  mean_score_time  std_score_time param_C  \\\n",
       "0       0.001624      0.000504         0.001539        0.000474       1   \n",
       "1       0.001491      0.000447         0.000802        0.000401       1   \n",
       "2       0.001172      0.000340         0.000935        0.000535      10   \n",
       "3       0.001204      0.000752         0.001028        0.000353      10   \n",
       "4       0.001119      0.000237         0.000891        0.000189      20   \n",
       "5       0.000907      0.000193         0.001001        0.000012      20   \n",
       "\n",
       "  param_kernel                         params  split0_test_score  \\\n",
       "0          rbf      {'C': 1, 'kernel': 'rbf'}           0.966667   \n",
       "1       linear   {'C': 1, 'kernel': 'linear'}           0.966667   \n",
       "2          rbf     {'C': 10, 'kernel': 'rbf'}           0.966667   \n",
       "3       linear  {'C': 10, 'kernel': 'linear'}           1.000000   \n",
       "4          rbf     {'C': 20, 'kernel': 'rbf'}           0.966667   \n",
       "5       linear  {'C': 20, 'kernel': 'linear'}           1.000000   \n",
       "\n",
       "   split1_test_score  split2_test_score  split3_test_score  split4_test_score  \\\n",
       "0                1.0           0.966667           0.966667                1.0   \n",
       "1                1.0           0.966667           0.966667                1.0   \n",
       "2                1.0           0.966667           0.966667                1.0   \n",
       "3                1.0           0.900000           0.966667                1.0   \n",
       "4                1.0           0.900000           0.966667                1.0   \n",
       "5                1.0           0.900000           0.933333                1.0   \n",
       "\n",
       "   mean_test_score  std_test_score  rank_test_score  \n",
       "0         0.980000        0.016330                1  \n",
       "1         0.980000        0.016330                1  \n",
       "2         0.980000        0.016330                1  \n",
       "3         0.973333        0.038873                4  \n",
       "4         0.966667        0.036515                5  \n",
       "5         0.966667        0.042164                6  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# last answer into pandas dataframe\n",
    "df = pd.DataFrame(clf.cv_results_)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "daf05170-054a-4825-8cfc-1febe5a32d94",
   "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>param_C</th>\n",
       "      <th>param_kernel</th>\n",
       "      <th>mean_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>rbf</td>\n",
       "      <td>0.980000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>linear</td>\n",
       "      <td>0.980000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>rbf</td>\n",
       "      <td>0.980000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>linear</td>\n",
       "      <td>0.973333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>rbf</td>\n",
       "      <td>0.966667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>20</td>\n",
       "      <td>linear</td>\n",
       "      <td>0.966667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  param_C param_kernel  mean_test_score\n",
       "0       1          rbf         0.980000\n",
       "1       1       linear         0.980000\n",
       "2      10          rbf         0.980000\n",
       "3      10       linear         0.973333\n",
       "4      20          rbf         0.966667\n",
       "5      20       linear         0.966667"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[['param_C','param_kernel','mean_test_score']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "bb2aa547-3100-42c7-8440-aed27c7e4cc2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['__abstractmethods__',\n",
       " '__annotations__',\n",
       " '__class__',\n",
       " '__delattr__',\n",
       " '__dict__',\n",
       " '__dir__',\n",
       " '__doc__',\n",
       " '__eq__',\n",
       " '__format__',\n",
       " '__ge__',\n",
       " '__getattribute__',\n",
       " '__getstate__',\n",
       " '__gt__',\n",
       " '__hash__',\n",
       " '__init__',\n",
       " '__init_subclass__',\n",
       " '__le__',\n",
       " '__lt__',\n",
       " '__module__',\n",
       " '__ne__',\n",
       " '__new__',\n",
       " '__reduce__',\n",
       " '__reduce_ex__',\n",
       " '__repr__',\n",
       " '__setattr__',\n",
       " '__setstate__',\n",
       " '__sizeof__',\n",
       " '__sklearn_clone__',\n",
       " '__str__',\n",
       " '__subclasshook__',\n",
       " '__weakref__',\n",
       " '_abc_impl',\n",
       " '_build_request_for_signature',\n",
       " '_check_feature_names',\n",
       " '_check_n_features',\n",
       " '_check_refit_for_multimetric',\n",
       " '_doc_link_module',\n",
       " '_doc_link_template',\n",
       " '_doc_link_url_param_generator',\n",
       " '_estimator_type',\n",
       " '_format_results',\n",
       " '_get_default_requests',\n",
       " '_get_doc_link',\n",
       " '_get_metadata_request',\n",
       " '_get_param_names',\n",
       " '_get_routed_params_for_fit',\n",
       " '_get_scorers',\n",
       " '_get_tags',\n",
       " '_more_tags',\n",
       " '_parameter_constraints',\n",
       " '_repr_html_',\n",
       " '_repr_html_inner',\n",
       " '_repr_mimebundle_',\n",
       " '_required_parameters',\n",
       " '_run_search',\n",
       " '_select_best_index',\n",
       " '_validate_data',\n",
       " '_validate_params',\n",
       " 'best_estimator_',\n",
       " 'best_index_',\n",
       " 'best_params_',\n",
       " 'best_score_',\n",
       " 'classes_',\n",
       " 'cv',\n",
       " 'cv_results_',\n",
       " 'decision_function',\n",
       " 'error_score',\n",
       " 'estimator',\n",
       " 'fit',\n",
       " 'get_metadata_routing',\n",
       " 'get_params',\n",
       " 'inverse_transform',\n",
       " 'multimetric_',\n",
       " 'n_features_in_',\n",
       " 'n_jobs',\n",
       " 'n_splits_',\n",
       " 'param_grid',\n",
       " 'pre_dispatch',\n",
       " 'predict',\n",
       " 'predict_log_proba',\n",
       " 'predict_proba',\n",
       " 'refit',\n",
       " 'refit_time_',\n",
       " 'return_train_score',\n",
       " 'score',\n",
       " 'score_samples',\n",
       " 'scorer_',\n",
       " 'scoring',\n",
       " 'set_params',\n",
       " 'transform',\n",
       " 'verbose']"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dir(clf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "553b5a13-975b-4901-8218-627f75c14650",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9800000000000001"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "d5ecef90-55dd-4080-89c0-09bd2edc2873",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 1, 'kernel': 'rbf'}"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.best_params_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5d77dd78-543b-4eb9-beb3-58a38325aa4e",
   "metadata": {},
   "source": [
    "# \n",
    "Use RandomizedSearchCV to reduce number of iterations and with random combination of parameters. This is useful when you have too many parameters to try and your training time is longer. It helps reduce the cost of computation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "644c9b6f-8406-4996-813f-dc1026b18eca",
   "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>param_C</th>\n",
       "      <th>param_kernel</th>\n",
       "      <th>mean_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>20</td>\n",
       "      <td>linear</td>\n",
       "      <td>0.966667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>linear</td>\n",
       "      <td>0.973333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  param_C param_kernel  mean_test_score\n",
       "0      20       linear         0.966667\n",
       "1      10       linear         0.973333"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "rs = RandomizedSearchCV(svm.SVC(gamma='auto'), {\n",
    "        'C': [1,10,20],\n",
    "        'kernel': ['rbf','linear']\n",
    "    }, \n",
    "    cv=5, \n",
    "    return_train_score=False, \n",
    "    n_iter=2\n",
    ")\n",
    "rs.fit(iris.data, iris.target)\n",
    "pd.DataFrame(rs.cv_results_)[['param_C','param_kernel','mean_test_score']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a74ad85-adf0-45d3-920c-85babeef5e95",
   "metadata": {},
   "source": [
    "# \n",
    "How about different models with different hyperparameters?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "06257a9e-fe96-455d-b1b5-066c104df501",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import svm\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "model_params = {\n",
    "    'svm': {\n",
    "        'model': svm.SVC(gamma='auto'),\n",
    "        'params' : {\n",
    "            'C': [1,10,20],\n",
    "            'kernel': ['rbf','linear']\n",
    "        }  \n",
    "    },\n",
    "    'random_forest': {\n",
    "        'model': RandomForestClassifier(),\n",
    "        'params' : {\n",
    "            'n_estimators': [1,5,10]\n",
    "        }\n",
    "    },\n",
    "    'logistic_regression' : {\n",
    "        'model': LogisticRegression(solver='liblinear',multi_class='auto'),\n",
    "        'params': {\n",
    "            'C': [1,5,10]\n",
    "        }\n",
    "    }\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "c8edced5-3658-4742-b530-aefd1f1d9491",
   "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>model</th>\n",
       "      <th>best_score</th>\n",
       "      <th>best_params</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>svm</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>{'C': 1, 'kernel': 'rbf'}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>random_forest</td>\n",
       "      <td>0.953333</td>\n",
       "      <td>{'n_estimators': 5}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>logistic_regression</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>{'C': 5}</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 model  best_score                best_params\n",
       "0                  svm    0.980000  {'C': 1, 'kernel': 'rbf'}\n",
       "1        random_forest    0.953333        {'n_estimators': 5}\n",
       "2  logistic_regression    0.966667                   {'C': 5}"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores = []\n",
    "\n",
    "for model_name, mp in model_params.items():\n",
    "    clf =  GridSearchCV(mp['model'], mp['params'], cv=5, return_train_score=False)\n",
    "    clf.fit(iris.data, iris.target)\n",
    "    scores.append({\n",
    "        'model': model_name,\n",
    "        'best_score': clf.best_score_,\n",
    "        'best_params': clf.best_params_\n",
    "    })\n",
    "    \n",
    "df = pd.DataFrame(scores,columns=['model','best_score','best_params'])\n",
    "df"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}