完成逻辑回归的python方法

本文实现的原理很简单，优化方法是用的梯度下降。后面有测试结果。

先来看看实现的示例代码：

# coding=utf-8
from math import exp

import matplotlib.pyplot as plt
import numpy as np
from sklearn.datasets.samples_generator import make_blobs


def sigmoid(num):
 &＃39;&＃39;&＃39;

 :param num: 待计算的x
 :return: sigmoid之后的数值
 &＃39;&＃39;&＃39;
 if type(num) == int or type(num) == float:
  return 1.0 / (1 + exp(-1 * num))
 else:
  raise ValueError, &＃39;only int or float data can compute sigmoid&＃39;


class logistic():
 def init(self, x, y): 
  if type(x) == type(y) == list:
   self.x = np.array(x)
   self.y = np.array(y)
  elif type(x) == type(y) == np.ndarray:
   self.x = x
   self.y = y
  else:
   raise ValueError, &＃39;input data error&＃39;

 def sigmoid(self, x):
  &＃39;&＃39;&＃39;

  :param x: 输入向量
  :return: 对输入向量整体进行simgoid计算后的向量结果
  &＃39;&＃39;&＃39;
  s = np.frompyfunc(lambda x: sigmoid(x), 1, 1)
  return s(x)

 def train_with_punish(self, alpha, errors, punish=0.0001):
  &＃39;&＃39;&＃39;

  :param alpha: alpha为学习速率
  :param errors: 误差小于多少时停止迭代的阈值
  :param punish: 惩罚系数
  :param times: 最大迭代次数
  :return:
  &＃39;&＃39;&＃39;
  self.punish = punish
  dimension = self.x.shape[1]
  self.theta = np.random.random(dimension)
  compute_error = 100000000
  times = 0
  while compute_error > errors:
   res = np.dot(self.x, self.theta)
   delta = self.sigmoid(res) - self.y
   self.theta = self.theta - alpha * np.dot(self.x.T, delta) - punish * self.theta # 带惩罚的梯度下降方法
   compute_error = np.sum(delta)
   times += 1

 def predict(self, x):
  &＃39;&＃39;&＃39;

  :param x: 给入新的未标注的向量
  :return: 按照计算出的参数返回判定的类别
  &＃39;&＃39;&＃39;
  x = np.array(x)
  if self.sigmoid(np.dot(x, self.theta)) > 0.5:
   return 1
  else:
   return 0


def test1():
 &＃39;&＃39;&＃39;
 用来进行测试和画图，展现效果
 :return:
 &＃39;&＃39;&＃39;
 x, y = make_blobs(n_samples=200, centers=2, n_features=2, random_state=0, center_box=(10, 20))
 x1 = []
 y1 = []
 x2 = []
 y2 = []
 for i in range(len(y)):
  if y[i] == 0:
   x1.append(x[i][0])
   y1.append(x[i][1])
  elif y[i] == 1:
   x2.append(x[i][0])
   y2.append(x[i][1])
 # 以上均为处理数据，生成出两类数据
 p = logistic(x, y)
 p.train_with_punish(alpha=0.00001, errors=0.005, punish=0.01) # 步长是0.00001，最大允许误差是0.005，惩罚系数是0.01
 x_test = np.arange(10, 20, 0.01)
 y_test = (-1 * p.theta[0] / p.theta[1]) * x_test
 plt.plot(x_test, y_test, c=&＃39;g&＃39;, label=&＃39;logistic_line&＃39;)
 plt.scatter(x1, y1, c=&＃39;r&＃39;, label=&＃39;positive&＃39;)
 plt.scatter(x2, y2, c=&＃39;b&＃39;, label=&＃39;negative&＃39;)
 plt.legend(loc=2)
 plt.title(&＃39;punish value = &＃39; + p.punish.str())
 plt.show()


if name == &＃39;main&＃39;:
 test1()

运行结果如下图

总结