参考网址：

http://www.cnblogs.com/begtostudy/archive/2010/08/03/1790935.html

b=np.arange(15)
print(b)

[ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14]

b.reshape(3,5)

Out[32]:
array([[ 0,  1,  2,  3,  4],
       [ 5,  6,  7,  8,  9],
       [10, 11, 12, 13, 14]])

c=np.zeros((4,5))
print (c)

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

d=np.ones((5,7))
print (d)

[[ 1.  1.  1.  1.  1.  1.  1.]
 [ 1.  1.  1.  1.  1.  1.  1.]
 [ 1.  1.  1.  1.  1.  1.  1.]
 [ 1.  1.  1.  1.  1.  1.  1.]
 [ 1.  1.  1.  1.  1.  1.  1.]]

e=np.add(c,np.arange(20).reshape(4,5))
print (e)

[[  0.   1.   2.   3.   4.]
 [  5.   6.   7.   8.   9.]
 [ 10.  11.  12.  13.  14.]
 [ 15.  16.  17.  18.  19.]]

f=np.dot(e,d)
print (f)

[[ 10.  10.  10.  10.  10.  10.  10.]
 [ 35.  35.  35.  35.  35.  35.  35.]
 [ 60.  60.  60.  60.  60.  60.  60.]
 [ 85.  85.  85.  85.  85.  85.  85.]]

如果还不明白，举个简单的列子

t1=np.ones((2,2))
t2=np.arange(4).reshape(2,2)
t3=np.dot(t1,t2)
print (t3)

[[ 2.  4.]
 [ 2.  4.]]

即

[[ 1*0+1*2.  1*1+1*3.]
 [ 2.  4.]]

numpy.linalg中有标准的矩阵运算