import numpy as np
import pandas as pd
from sklearn.datasets import load_breast_cancer
from sklearn.model_selection import train_test_split

#데이터 가져옴
data = load_breast_cancer()  
#데이터 키로 설명변수
X = data['data']  
# Relabel such that 0 = 'benign' and 1 = malignant.
Y = 1 - data['target']  #타겟키로 반응변수  0과 1을 역전시킴

#데이터셋을 train, test로 쪼갬
X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.4, random_state=1234)

#공식 사용
def sigmoid(x):
    s = 1.0/(1.0 + np.exp(-x))  
    return s

def gradient(X, Y, beta):
    z = np.dot(X,beta.T)*Y
    ds = -Y*(1-sigmoid(z))*X
    return ds.sum(axis=0)

class LogisticRegression:
    def __init__(self, learn_rate):
        self.rate = learn_rate
        self.n_nodes = None
        self.beta = None
        
    def train(self, input_X, input_Y, n_epochs):
        self.n_nodes = input_X.shape[1] + 1
        self.beta = np.random.normal(0.0,1.0,(1,self.n_nodes))
        ones_column = np.ones((input_X.shape[0],1))
        X = np.concatenate((ones_column,input_X),axis=1)
        Y = (2*input_Y - 1).reshape(-1,1)
        for n in range(n_epochs):
            self.beta = self.beta - self.rate*gradient(X,Y,self.beta)
        return self.beta
    
    def query(self, input_X, prob=True, cutoff=0.5):
        ones_column = np.ones((input_X.shape[0],1))
        X = np.concatenate((ones_column,input_X),axis=1)
        z = np.dot(X,(self.beta).T)
        p = sigmoid(z)
        if prob :
            return p
        else:
            return (p > cutoff).astype('int')

# Hyperparameter for the learner.
learning_rate = 0.001

# Train and predict.
LR = LogisticRegression(learning_rate)
LR.train(X_train, Y_train, 2000)
Y_pred = LR.query(X_test,prob=False,cutoff=0.5)