TypeError: only length-1 arrays can be converted to Python scalars

def sigmoid(intX):
    return 1.0/(1+exp(-intX))
dataMatrix = mat(dataMatIn)
weights = ones((n, 1))
h = sigmoid(dataMatrix*weights)

出错：

return 1.0/(1+math.exp(-intX)) 
TypeError: only length-1 arrays can be converted to Python scalars

因为dataMatrix 与weights均为numpy矩阵，相乘也是numpy矩阵，而math.exp()函数只处理python标准数值。
此处需要用numpy的exp()方法，如下：

import numpy as np
def sigmoid(self, intX):
    return 1.0/(1+np.exp(-intX))

也可以在文件头添加from numpy import *,就可以直接用exp(-intX)了

---

def smoSimple(dataMatIn, classLabels, C, toler, maxIter):
    dataMatrix = mat(dataMatIn)
    labelMat = mat(classLabels).transpose()
    iter = 0
    while iter < maxIter:
        alphaPairChanged = 0
        for i in range(m):
            fXi = float(multiply(alphas, labelMat).T * (dataMatrix * dataMatrix[i,:].T)) + b
           .....

出错：

    fXi = float(multiply(alphas, labelMat).T * (dataMatrix * dataMatrix[i,:].T)) + b
TypeError: only length-1 arrays can be converted to Python scalars

print 
multiply(alphas, labelMat).T * (dataMatrix * dataMatrix[i,:].T)

[[ 0.]
 [ 0.]
 [ 0.]
 [ 0.]
 ...
 [ 0.]]

可见float()函数中是一个numpy数组，此例又证明标准python函数对numpy数组不适用。