numpy中的axis怎么理解：数组的维数，比如4*5数组，行为axis＝1，列为axis＝0
eg：
a = np.arange(20).reshape(4,5)
print "a:"
print a
print "maximum element in each row of a: " + str(a.max(axis=1))
print "minimum element in each column of a: " + str(a.min(axis=0))
原文：http://www.cnblogs.com/amanon/p/6349251.html