ROC_OptimumThreshold.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9543af60",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Continuation of the previous problem named as ConfusionMatrix_ROCcurve_logistic_regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "78b77cf2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1519e785",
   "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>Customer</th>\n",
       "      <th>Spending</th>\n",
       "      <th>Card</th>\n",
       "      <th>Coupon</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2.291</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>3.215</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>2.135</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>3.924</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>2.528</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>96</td>\n",
       "      <td>3.318</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>97</td>\n",
       "      <td>2.421</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>98</td>\n",
       "      <td>6.073</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>99</td>\n",
       "      <td>2.630</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>100</td>\n",
       "      <td>3.411</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    Customer  Spending  Card  Coupon\n",
       "0          1     2.291     1       0\n",
       "1          2     3.215     1       0\n",
       "2          3     2.135     1       0\n",
       "3          4     3.924     0       0\n",
       "4          5     2.528     1       0\n",
       "..       ...       ...   ...     ...\n",
       "95        96     3.318     0       0\n",
       "96        97     2.421     1       0\n",
       "97        98     6.073     0       0\n",
       "98        99     2.630     1       0\n",
       "99       100     3.411     0       1\n",
       "\n",
       "[100 rows x 4 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_excel(\"/home/ashish/projects/ROC_Curve_AND_Optimal_Threshold/Simmons2-Copy1.xls\")\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "490b12dc",
   "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>Customer</th>\n",
       "      <th>Spending</th>\n",
       "      <th>Card</th>\n",
       "      <th>Coupon</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>3.333790</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>29.011492</td>\n",
       "      <td>1.741298</td>\n",
       "      <td>0.502519</td>\n",
       "      <td>0.492366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.058000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.750000</td>\n",
       "      <td>2.059000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>2.805500</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.250000</td>\n",
       "      <td>4.468250</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>7.076000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Customer    Spending        Card      Coupon\n",
       "count  100.000000  100.000000  100.000000  100.000000\n",
       "mean    50.500000    3.333790    0.500000    0.400000\n",
       "std     29.011492    1.741298    0.502519    0.492366\n",
       "min      1.000000    1.058000    0.000000    0.000000\n",
       "25%     25.750000    2.059000    0.000000    0.000000\n",
       "50%     50.500000    2.805500    0.500000    0.000000\n",
       "75%     75.250000    4.468250    1.000000    1.000000\n",
       "max    100.000000    7.076000    1.000000    1.000000"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()  ## t describe the statistical information about data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c98023b7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Coupon'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "760b60b7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    60\n",
       "1    40\n",
       "Name: Coupon, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Coupon'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d5020526",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "60/100      ## using the baseline for the threshold"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2cf9bd9",
   "metadata": {},
   "source": [
    "## Building the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "b7c7162b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import linear_model\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5556db11",
   "metadata": {},
   "outputs": [],
   "source": [
    "## splitting the dataset into training and testing dataset\n",
    "x = data[['Card','Spending']]  ## independent variables\n",
    "y = data['Coupon'].values.reshape(-1,1)  ## dependent variable - 100 rows and 1 column\n",
    "x_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.25,random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "4858b9d9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(75, 75, 25, 25)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x_train),len(y_train),len(x_test),len(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "f19a1089",
   "metadata": {},
   "outputs": [],
   "source": [
    "## this below method is used to chek the available methods and parameters \n",
    "##LogisticRegression?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7424f497",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression()"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Logisticreg = LogisticRegression(solver='lbfgs')\n",
    "Logisticreg.fit(x_train,y_train.ravel())  ## ravel will return 1D array with all the input array elements"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1190ae4e",
   "metadata": {},
   "source": [
    "## Predict the values of dependent variable using test and training set of independent variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d3d3a21b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,\n",
       "       1, 0, 0])"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict_test = Logisticreg.predict(x_test)\n",
    "y_predict_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "79266920",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 1,\n",
       "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "       0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "       1, 1, 0, 0, 0, 0, 1, 1, 0])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict_train = Logisticreg.predict(x_train)\n",
    "y_predict_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "548edc93",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.49622117, 0.32880793, 0.44329114, 0.33320924, 0.41456465,\n",
       "        0.32890329, 0.3975043 , 0.66921229, 0.25844531, 0.63672372,\n",
       "        0.29274386, 0.28466974, 0.5159296 , 0.41992276, 0.24342356,\n",
       "        0.528514  , 0.47965107, 0.52805789, 0.33191449, 0.27457435,\n",
       "        0.49179296, 0.63261616, 0.24690181, 0.47089452, 0.27842076,\n",
       "        0.41663875, 0.36155602, 0.49970327, 0.23621636, 0.37860052,\n",
       "        0.48809323, 0.28877877, 0.28563859, 0.37231882, 0.65309742,\n",
       "        0.43807264, 0.33638478, 0.40406607, 0.23431177, 0.37282384,\n",
       "        0.49970327, 0.39768396, 0.32880793, 0.25782472, 0.47393834,\n",
       "        0.42878861, 0.26520939, 0.33320924, 0.54682499, 0.45446086,\n",
       "        0.44326597, 0.4965167 , 0.60065954, 0.38989654, 0.49149447,\n",
       "        0.27414424, 0.27785686, 0.67464141, 0.28195004, 0.48593427,\n",
       "        0.38633222, 0.31373499, 0.42810085, 0.27418723, 0.44371771,\n",
       "        0.41629601, 0.642004  , 0.6571001 , 0.44068025, 0.28195004,\n",
       "        0.40217015, 0.43807264, 0.50977653, 0.57944626, 0.2904233 ]])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Calculating the probabilities\n",
    "y_prob_train = Logisticreg.predict_proba(x_train)[:,1]\n",
    "##print(y_prob_train)\n",
    "y_prob_train.reshape(1,-1)   ## 75 rows and 1 column"
   ]
  },
  {
   "attachments": {
    "predict_proba.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "1d3a76fa",
   "metadata": {},
   "source": [
    "![predict_proba.png](attachment:predict_proba.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "99e1c650",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.52802946, 0.49516653, 0.45703306, 0.27712052, 0.34390047,\n",
       "        0.26825171, 0.27712052, 0.607686  , 0.42836534, 0.43637155,\n",
       "        0.31387455, 0.23676248, 0.45703306, 0.43602768, 0.37596116,\n",
       "        0.44900317, 0.46952365, 0.68521935, 0.25167254, 0.47073304,\n",
       "        0.42361093, 0.56580644, 0.52792177, 0.40302605, 0.27457435]])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_prob_test = Logisticreg.predict_proba(x_test)[:,1]\n",
    "##print(y_prob_test)\n",
    "y_prob_test.reshape(1,-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "c7272915",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.604869\n",
      "         Iterations 5\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Logit Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>        <td>Coupon</td>      <th>  No. Observations:  </th>  <td>   100</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                 <td>Logit</td>      <th>  Df Residuals:      </th>  <td>    97</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>                 <td>MLE</td>       <th>  Df Model:          </th>  <td>     2</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>            <td>Sat, 29 Jan 2022</td> <th>  Pseudo R-squ.:     </th>  <td>0.1012</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                <td>23:01:49</td>     <th>  Log-Likelihood:    </th> <td> -60.487</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>converged:</th>             <td>True</td>       <th>  LL-Null:           </th> <td> -67.301</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>     <td>nonrobust</td>    <th>  LLR p-value:       </th> <td>0.001098</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>        <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>const</th>    <td>   -2.1464</td> <td>    0.577</td> <td>   -3.718</td> <td> 0.000</td> <td>   -3.278</td> <td>   -1.015</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Spending</th> <td>    0.3416</td> <td>    0.129</td> <td>    2.655</td> <td> 0.008</td> <td>    0.089</td> <td>    0.594</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Card</th>     <td>    1.0987</td> <td>    0.445</td> <td>    2.471</td> <td> 0.013</td> <td>    0.227</td> <td>    1.970</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                           Logit Regression Results                           \n",
       "==============================================================================\n",
       "Dep. Variable:                 Coupon   No. Observations:                  100\n",
       "Model:                          Logit   Df Residuals:                       97\n",
       "Method:                           MLE   Df Model:                            2\n",
       "Date:                Sat, 29 Jan 2022   Pseudo R-squ.:                  0.1012\n",
       "Time:                        23:01:49   Log-Likelihood:                -60.487\n",
       "converged:                       True   LL-Null:                       -67.301\n",
       "Covariance Type:            nonrobust   LLR p-value:                  0.001098\n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "const         -2.1464      0.577     -3.718      0.000      -3.278      -1.015\n",
       "Spending       0.3416      0.129      2.655      0.008       0.089       0.594\n",
       "Card           1.0987      0.445      2.471      0.013       0.227       1.970\n",
       "==============================================================================\n",
       "\"\"\""
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = data[['Spending','Card']]\n",
    "y = data['Coupon']\n",
    "\n",
    "import statsmodels.api as sm\n",
    "x1 = sm.add_constant(x)\n",
    "logit_model = sm.Logit(y,x1).fit()\n",
    "logit_model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4cc2bd3",
   "metadata": {},
   "source": [
    "## Logistic Regression Equation coefficients::\n",
    "    spending --> 0.3416\n",
    "    Card --> 1.0987\n",
    "    const = -2.1464"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ec87798",
   "metadata": {},
   "source": [
    "# ACCURACY SCORE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "0f6c8eb0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.76"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Checking the Accuracy of our model\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "score =  accuracy_score(y_test,y_predict_test)\n",
    "score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "209d105d",
   "metadata": {},
   "source": [
    "# CONFUSION MATRIX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "71fc12f7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[15,  1],\n",
       "       [ 5,  4]])"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Checking the confusion matrix of our model\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "confusion_matrix(y_test,y_predict_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "6cc2dd3b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Negatives ::  15\n",
      "False Positives ::  1\n",
      "True Positives ::  4\n",
      "False Negatives ::  5\n"
     ]
    }
   ],
   "source": [
    "tn,fp,fn,tp = confusion_matrix(y_test,y_predict_test).ravel()\n",
    "print(\"True Negatives :: \",tn)\n",
    "print(\"False Positives :: \",fp)\n",
    "print(\"True Positives :: \",tp)\n",
    "print(\"False Negatives :: \",fn)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc11429c",
   "metadata": {},
   "source": [
    "# CLASSIFICATION REPORT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "88dc5300",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.75      0.94      0.83        16\n",
      "           1       0.80      0.44      0.57         9\n",
      "\n",
      "    accuracy                           0.76        25\n",
      "   macro avg       0.78      0.69      0.70        25\n",
      "weighted avg       0.77      0.76      0.74        25\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test,y_predict_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "7841d102",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.76\n"
     ]
    }
   ],
   "source": [
    "Accuracy = (tp+tn)/(tp+tn+fp+fn)\n",
    "print(\"Accuracy {:0.2f}\".format(Accuracy))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "ed4f7c6b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Specificty 0.94\n"
     ]
    }
   ],
   "source": [
    "Specificity = tn/(tn+fp)\n",
    "print(\"Specificty {:0.2f}\".format(Specificity))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "9dd691bb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sensitivity 0.44\n"
     ]
    }
   ],
   "source": [
    "Sensitivity = tp/(tp+fn)\n",
    "print(\"Sensitivity {:0.2f}\".format(Sensitivity))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "78d34248",
   "metadata": {},
   "source": [
    "# ROC CURVE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "f43c6479",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve,auc\n",
    "\n",
    "log_ROC_AUC1 = roc_auc_score(y_train,y_predict_train)\n",
    "fpr1,tpr1,threshold1 = roc_curve(y_train,y_prob_train)\n",
    "roc_auc1 = auc(fpr1,tpr1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "5d11dbc2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.67464141 0.67464141 0.6571001  0.65309742 0.528514   0.50977653\n",
      " 0.49970327 0.4965167  0.49622117 0.48809323 0.44068025 0.43807264\n",
      " 0.41992276 0.41629601 0.40406607 0.40217015 0.38633222 0.37282384\n",
      " 0.36155602 0.33638478 0.33320924 0.33191449 0.32890329 0.32880793\n",
      " 0.31373499 0.29274386 0.28877877 0.28563859 0.28466974 0.28195004\n",
      " 0.27457435 0.25844531 0.25782472 0.23431177] \n",
      "\n",
      "\n",
      "\n",
      "[0.         0.         0.         0.02272727 0.02272727 0.09090909\n",
      " 0.11363636 0.11363636 0.13636364 0.13636364 0.34090909 0.36363636\n",
      " 0.43181818 0.43181818 0.47727273 0.47727273 0.56818182 0.56818182\n",
      " 0.61363636 0.61363636 0.63636364 0.65909091 0.65909091 0.70454545\n",
      " 0.72727273 0.72727273 0.77272727 0.77272727 0.79545455 0.81818182\n",
      " 0.81818182 0.90909091 0.90909091 1.        ] \n",
      "\n",
      "\n",
      "\n",
      "\n",
      "[0.         0.03225806 0.09677419 0.09677419 0.32258065 0.32258065\n",
      " 0.35483871 0.38709677 0.38709677 0.48387097 0.48387097 0.51612903\n",
      " 0.51612903 0.58064516 0.58064516 0.61290323 0.61290323 0.67741935\n",
      " 0.67741935 0.70967742 0.74193548 0.74193548 0.77419355 0.77419355\n",
      " 0.77419355 0.80645161 0.80645161 0.83870968 0.83870968 0.87096774\n",
      " 0.96774194 0.96774194 1.         1.        ]\n"
     ]
    }
   ],
   "source": [
    "print(threshold1,\"\\n\\n\\n\")\n",
    "print(fpr1,\"\\n\\n\\n\\n\")\n",
    "print(tpr1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "30d178fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6158357771260996\n"
     ]
    }
   ],
   "source": [
    "print(log_ROC_AUC1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "6c741cdc",
   "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": [
    "plt.figure()\n",
    "plt.plot(fpr1,tpr1,color='blue',label='ROC curve (area = %0.2f)'%roc_auc1)\n",
    "plt.plot([0,1],[0,1],'r--')\n",
    "plt.xlim([0.0,1.0])\n",
    "plt.ylim([0.0,1.05])\n",
    "plt.xlabel(\"FALSE POSITIVE RATE\")\n",
    "plt.ylabel(\"TRUE POSITIVE RATE\")\n",
    "plt.title(\"Reciever operating characteristics example\")\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "740e83aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "log_ROC_AUC = roc_auc_score(y_test,y_predict_test)\n",
    "fpr,tpr,threshold = roc_curve(y_test,y_prob_test)\n",
    "roc_auc = auc(fpr,tpr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "9fd2e104",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.     0.     0.     0.0625 0.0625 0.125  0.125  0.125  0.375  0.375\n",
      " 0.625  0.75   1.    ] \n",
      "\n",
      "\n",
      "\n",
      "\n",
      "[0.         0.11111111 0.33333333 0.33333333 0.55555556 0.55555556\n",
      " 0.66666667 0.88888889 0.88888889 1.         1.         1.\n",
      " 1.        ]\n"
     ]
    }
   ],
   "source": [
    "print(fpr,\"\\n\\n\\n\\n\")\n",
    "print(tpr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "2230a720",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f2781d60460>"
      ]
     },
     "execution_count": 82,
     "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": [
    "plt.figure()\n",
    "plt.plot(fpr,tpr,color=\"blue\",label=\"ROC curve (area = %0.2f)\"%roc_auc)\n",
    "plt.plot([0,1],[0,1],'r--')\n",
    "plt.xlim(0,1)\n",
    "plt.ylim(0,1.05)\n",
    "plt.title(\"Receiver Operating Characteristic Example\")\n",
    "plt.legend(loc=\"lower right\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4ce7c6c",
   "metadata": {},
   "source": [
    "# first case: set threshold =0.35 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "b95f0265",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 1., 1., 0., 0., 0., 0., 1., 1., 1., 0., 0., 1., 1., 1., 1., 1.,\n",
       "       1., 0., 1., 1., 1., 1., 1., 0.])"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import binarize\n",
    "y_predict_class1 = binarize(y_prob_test.reshape(1,-1),threshold=0.35)[0]\n",
    "y_predict_class1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "afe8f578",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1,\n",
       "       1, 1, 0])"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict_class1 = y_predict_class1.astype(int)\n",
    "y_predict_class1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "87534350",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[8, 8],\n",
       "       [0, 9]])"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix_2 = confusion_matrix(y_test,y_predict_class1)\n",
    "confusion_matrix_2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "d4b81e1f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True negative :  8\n",
      "False positive :  8\n",
      "False negative :  0\n",
      "True Positive :  9\n"
     ]
    }
   ],
   "source": [
    "tn,fp,fn,tp = confusion_matrix(y_test,y_predict_class1).ravel()\n",
    "print(\"True negative : \",tn)\n",
    "print(\"False positive : \",fp)\n",
    "print(\"False negative : \",fn)\n",
    "print(\"True Positive : \",tp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "65634857",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.50      0.67        16\n",
      "           1       0.53      1.00      0.69         9\n",
      "\n",
      "    accuracy                           0.68        25\n",
      "   macro avg       0.76      0.75      0.68        25\n",
      "weighted avg       0.83      0.68      0.68        25\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test,y_predict_class1))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e2fc9fd6",
   "metadata": {},
   "source": [
    "# Second Case: set threshold = 0.50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "c6dda6c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       1., 0., 0., 0., 1., 1., 0., 0.])"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import binarize\n",
    "y_predict_class2 = binarize(y_prob_test.reshape(1,-1),threshold=0.50)[0]\n",
    "y_predict_class2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "7336a2d0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,\n",
       "       1, 0, 0])"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict_class2 = y_predict_class2.astype(int)\n",
    "y_predict_class2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "eb1ee70d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[15  1]\n",
      " [ 5  4]]\n"
     ]
    }
   ],
   "source": [
    "confusion_matrix_3 = confusion_matrix(y_test,y_predict_class2)\n",
    "print(confusion_matrix_3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "bad0fa9b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True negative :  15\n",
      "False positive :  1\n",
      "False negative :  5\n",
      "True Positive :  4\n"
     ]
    }
   ],
   "source": [
    "tn,fp,fn,tp = confusion_matrix(y_test,y_predict_class2).ravel()\n",
    "print(\"True negative : \",tn)\n",
    "print(\"False positive : \",fp)\n",
    "print(\"False negative : \",fn)\n",
    "print(\"True Positive : \",tp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "f1ae09f7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.75      0.94      0.83        16\n",
      "           1       0.80      0.44      0.57         9\n",
      "\n",
      "    accuracy                           0.76        25\n",
      "   macro avg       0.78      0.69      0.70        25\n",
      "weighted avg       0.77      0.76      0.74        25\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test,y_predict_class2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e95ff51",
   "metadata": {},
   "source": [
    "# Third Case : --> Set threshold = 0.70"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "9baa9455",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0.])"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import binarize\n",
    "y_predict_class3 = binarize(y_prob_test.reshape(1,-1),threshold=0.70)[0]\n",
    "y_predict_class3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "8e499460",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "       0, 0, 0])"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict_class3 = y_predict_class3.astype(int)\n",
    "y_predict_class3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "e5f2d22b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[16  0]\n",
      " [ 9  0]]\n"
     ]
    }
   ],
   "source": [
    "confusion_matrix_4 = confusion_matrix(y_test,y_predict_class3) \n",
    "print(confusion_matrix_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "986b41b9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.64      1.00      0.78        16\n",
      "           1       0.00      0.00      0.00         9\n",
      "\n",
      "    accuracy                           0.64        25\n",
      "   macro avg       0.32      0.50      0.39        25\n",
      "weighted avg       0.41      0.64      0.50        25\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ashish/anaconda3/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/ashish/anaconda3/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/ashish/anaconda3/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test,y_predict_class3))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "570dd771",
   "metadata": {},
   "source": [
    "# FINAL STEP : --> FINDING OPTIMAL CUTOFF POINT OR THRESHOLD VALUE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "01de08b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve,auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "ef53c1fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "fpr,tpr,thresholds = roc_curve(y_test,y_prob_test) \n",
    "roc_auc = auc(fpr,tpr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "6313ec04",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under the ROC curve : 0.902778 \n"
     ]
    }
   ],
   "source": [
    "print(\"Area under the ROC curve : %f \"%roc_auc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "8895d950",
   "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>fpr</th>\n",
       "      <th>tpr</th>\n",
       "      <th>1-fpr</th>\n",
       "      <th>tf</th>\n",
       "      <th>thresholds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.125</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.013889</td>\n",
       "      <td>0.457033</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     fpr       tpr  1-fpr        tf  thresholds\n",
       "7  0.125  0.888889  0.875  0.013889    0.457033"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "i = np.arange(len(tpr))  ## index for df\n",
    "roc = pd.DataFrame({'fpr':pd.Series(fpr,index=i),\n",
    "                   'tpr':pd.Series(tpr,index=i),\n",
    "                   '1-fpr': pd.Series(1-fpr,index=i),\n",
    "                    'tf' : pd.Series(tpr-(1-fpr),index=i),\n",
    "                    'thresholds' : pd.Series(thresholds,index=i)\n",
    "                   })\n",
    "roc.iloc[(roc.tf-0).abs().argsort()[:1]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "dc1d78ad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(-2.0, 0, ''),\n",
       " Text(0.0, 0, ''),\n",
       " Text(2.0, 0, ''),\n",
       " Text(4.0, 0, ''),\n",
       " Text(6.0, 0, ''),\n",
       " Text(8.0, 0, ''),\n",
       " Text(10.0, 0, ''),\n",
       " Text(12.0, 0, ''),\n",
       " Text(14.0, 0, '')]"
      ]
     },
     "execution_count": 120,
     "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": [
    "fig,ax = plt.subplots()\n",
    "plt.plot(roc['tpr'])\n",
    "plt.plot(roc['1-fpr'],color='red')\n",
    "plt.xlabel('1-False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver Operating Charcateristics')\n",
    "ax.set_xticklabels([])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "b403c377",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 1., 1., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0., 1.,\n",
       "       1., 0., 1., 0., 1., 1., 0., 0.])"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import binarize\n",
    "y_predicted_class4 = binarize(y_prob_test.reshape(1,-1),threshold=0.45)[0]\n",
    "y_predicted_class4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "e98a95e7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 1, 0, 1,\n",
       "       1, 0, 0])"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predicted_class4 = y_predicted_class4.astype(int)\n",
    "y_predicted_class4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "2c94bc50",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[14  2]\n",
      " [ 1  8]]\n"
     ]
    }
   ],
   "source": [
    "confusion_matrix_5 = confusion_matrix(y_test,y_predicted_class4)\n",
    "print(confusion_matrix_5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "e4b2daa5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.88      0.90        16\n",
      "           1       0.80      0.89      0.84         9\n",
      "\n",
      "    accuracy                           0.88        25\n",
      "   macro avg       0.87      0.88      0.87        25\n",
      "weighted avg       0.89      0.88      0.88        25\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test,y_predicted_class4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "321ce548",
   "metadata": {},
   "source": [
    "# hence , at last we obtained the most optimum threshold for our logistic model."
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}