机器学习实战源码-----用线性回归找到最佳拟合曲线

 1 #_*_coding:utf-8_*_
 2 import numpy as np
 3 import matplotlib.pyplot as plt
 4 
 5 def loadDataSet(fileName):
 6     #numFeat表示特征的数量
 7     numFeat = len(open(fileName).readline().split("	")) - 1
 8     dataMat = [];labelMat = []
 9     fr = open(fileName)
10     for line in fr.readlines():
11         lineArr = []
12         curLine = line.strip().split("	")
13         for i in range(numFeat):
14             lineArr.append(float(curLine[i]))
15         dataMat.append(lineArr)
16         labelMat.append(float(curLine[-1]))
17     return dataMat,labelMat
18 
19 def standRegres(xArr,yArr):
20     xMat = np.mat(xArr);yMat = np.mat(yArr).T
21     xTx = xMat.T * xMat
22     if np.linalg.det(xTx) == 0.0:
23         print "This matrix is singular, cannot do inverse"
24         return
25     ws = xTx.I * (xMat.T * yMat)
26     return ws
27 
28 def lwlr(testPoint,xArr,yArr,k = 1.0):
29     xMat = np.mat(xArr);yMat = np.mat(yArr).T
30     m = np.shape(xMat)[0]
31     weights = np.mat(np.eye(m))
32     for j in range(m):
33         diffMat = testPoint - xMat[j,:]
34         weights[j,j] = np.exp(diffMat * diffMat.T / (-2.0 * k ** 2))
35     xTx = xMat.T * (weights * xMat)
36     if np.linalg.det(xTx) == 0.0:
37         print "This matrix is singular, cannot do inverse"
38         return
39     ws = xTx.I * (xMat.T * (weights * yMat))
40     return testPoint * ws
41 
42 def lwlrTest(testArr,xArr,yArr,k = 1.0):
43     m = np.shape(testArr)[0]
44     yHat = np.zeros(m)
45     for i in range(m):
46         yHat[i] = lwlr(testArr[i],xArr,yArr,k)
47     return yHat
48 
49 if __name__ == "__main__":
50     xArr,yArr = loadDataSet("ex0.txt")
51     yHat = lwlrTest(xArr,xArr,yArr,0.01)
52     xMat = np.mat(xArr)
53     srtInd = xMat[:,1].argsort(0)
54     xSort = xMat[srtInd][:,0,:]
55 
56     fig = plt.figure()
57     ax = fig.add_subplot(111)
58     ax.plot(xSort[:,1],yHat[srtInd])
59     ax.scatter(xMat[:,1].flatten().A[0],np.mat(yArr).T.flatten().A[0],s=2,c="red")
60     plt.show()