{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 로지스틱 회귀와 그래디언트 하강법"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.datasets import load_breast_cancer\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Read in data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#데이터 가져옴\n",
    "data = load_breast_cancer()  \n",
    "#데이터 키로 설명변수\n",
    "X = data['data']  \n",
    "# Relabel such that 0 = 'benign' and 1 = malignant.\n",
    "Y = 1 - data['target']  #타겟키로 반응변수  0과 1을 역전시킴"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#데이터셋을 train, test로 쪼갬\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.4, random_state=1234)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) 'sigmoid'및 'gradient'함수를 정의하여 아래 표시된 출력을 생성하라"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#공식 사용\n",
    "def sigmoid(x):\n",
    "    s = 1.0/(1.0 + np.exp(-x))  \n",
    "    return s\n",
    "\n",
    "def gradient(X, Y, beta):\n",
    "    z = np.dot(X,beta.T)*Y\n",
    "    ds = -Y*(1-sigmoid(z))*X\n",
    "    return ds.sum(axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) 'LogisticRegression'클래스를 정의하여 아래 표시된 출력을 생성하라"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "class LogisticRegression:\n",
    "    def __init__(self, learn_rate):\n",
    "        self.rate = learn_rate\n",
    "        self.n_nodes = None\n",
    "        self.beta = None\n",
    "        \n",
    "    def train(self, input_X, input_Y, n_epochs):\n",
    "        self.n_nodes = input_X.shape[1] + 1\n",
    "        self.beta = np.random.normal(0.0,1.0,(1,self.n_nodes))\n",
    "        ones_column = np.ones((input_X.shape[0],1))\n",
    "        X = np.concatenate((ones_column,input_X),axis=1)\n",
    "        Y = (2*input_Y - 1).reshape(-1,1)\n",
    "        for n in range(n_epochs):\n",
    "            self.beta = self.beta - self.rate*gradient(X,Y,self.beta)\n",
    "        return self.beta\n",
    "    \n",
    "    def query(self, input_X, prob=True, cutoff=0.5):\n",
    "        ones_column = np.ones((input_X.shape[0],1))\n",
    "        X = np.concatenate((ones_column,input_X),axis=1)\n",
    "        z = np.dot(X,(self.beta).T)\n",
    "        p = sigmoid(z)\n",
    "        if prob :\n",
    "            return p\n",
    "        else:\n",
    "            return (p > cutoff).astype('int')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3) Sample run"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Hyperparameter for the learner.\n",
    "learning_rate = 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train and predict.\n",
    "LR = LogisticRegression(learning_rate)\n",
    "LR.train(X_train, Y_train, 2000)\n",
    "Y_pred = LR.query(X_test,prob=False,cutoff=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy : 0.912\n"
     ]
    }
   ],
   "source": [
    "# Display the accuracy.\n",
    "acc = (Y_pred == Y_test.reshape(-1,1)).mean()  #예측된 y와 실제 y가 같느냐?하면 T/F로 나옴(논리배열 만들어짐)\n",
    "                                               #논리배열.mean을 하면 T는 1로, F는 0으로 계산\n",
    "print('Accuracy : {}'.format(np.round(acc,3)))  \n",
    "\n",
    "#정답의 비율이 바로 정확도 91.2%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}