逻辑回归实验

import random

from numpy import mat, array, arange
from numpy.ma import exp, ones, shape
import matplotlib.pyplot as plt

plt.ion()

def getData():
    fun = lambda x: 2 * x + 2

    # al = [fun(i,i+2) for i in range(100)]

    res = []
    labels = []

    for x in [random.random() * 5 for _ in range(100)]:
        for y in [random.random() * 20 for _ in range(100)]:
            res.append([x, y-2*x])
            if fun(x) > y:
                labels.append(1)
            else:
                labels.append(0)

    return res, labels


def sigmoid(inX):
    return 1.0 / (1 + exp(-inX))


def gradAscent(dataMatIn, classLabels):
    dataMatrix = mat(dataMatIn)  # 转换为 NumPy 矩阵
    labelMat = mat(classLabels).transpose()  # 首先将数组转换为 NumPy 矩阵，然后再将行向量转置为列向量
    m, n = shape(dataMatrix)
    alpha = 0.0001
    maxCycles = 100000
    weights = ones((n, 1))
    for k in range(maxCycles):
        h = sigmoid(dataMatrix * weights)

        error = (labelMat - h)

        # print(dataMatrix.transpose())
        # print(error)
        # 2行n列 * n行1列
        weights = weights + alpha * dataMatrix.transpose() * error

        if k % 1000:
            plotBestFit(dataMatIn, labelMat, weights)

    return array(weights)


def plotBestFit(dataArr, labelMat, weights):
    plt.clf()
    n = shape(dataArr)[0]
    xcord1 = []
    ycord1 = []
    xcord2 = []
    ycord2 = []
    for i in range(n):
        if int(labelMat[i]) == 1:
            xcord1.append(dataArr[i, 0])
            ycord1.append(dataArr[i, 1])
        else:
            xcord2.append(dataArr[i, 0])
            ycord2.append(dataArr[i, 1])
    # fig = plt.figure()
    # ax = fig.add_subplot(111)
    plt.scatter(xcord1, ycord1, s=3, c='red', marker='s')
    plt.scatter(xcord2, ycord2, s=3, c='green')
    x = arange(0, 5.0, 0.1)

    y = weights[0] * x + weights[1]

    plt.plot(x, y)
    plt.xlabel('X')
    plt.ylabel('Y')
    # plt.show()
    plt.pause(0.033)

def testLR():
    dataMat, labelMat = getData()
    dataArr = array(dataMat)
    weights = gradAscent(dataArr, labelMat)
    print(weights)



testLR()