logistic_regression_gd.py

# Algoritmo de gradient descent
class LogisticRegressionGD(object):
    
    def __init__(self, l_rate = 0.1, n_iter =10000, random_state =1):
        self.l_rate = l_rate
        self.n_iter = n_iter
        self.random_state = random_state               
        
    def fit(self, X, y):
        rgen = np.random.RandomState(self.random_state)        
        self.theta = rgen.normal(loc = 0.0, scale = 0.01,
                                 size = 1 + X.shape[1])     
        for i in range(self.n_iter):
            net_input = self.net_input(X)
            h = self.sigmoid(net_input)   
            errors = y-h
            self.theta[1:] += -self.l_rate*X.T.dot(errors) 
            self.theta[0] += -self.l_rate*errors.sum()          
        return self.theta
    
    def sigmoid(self, z):
        return 1. / (1. + np.exp(-np.clip(z, -250, 250)))
    
    def net_input(self, X):
        return np.dot(X, self.theta[1:]) + self.theta[0]
    
    def predict(self, X):        
        return np.where(self.sigmoid(self.net_input(X))>= 0.5, 0, 1)