Select rows in a Numpy 2D array with a boolean vector

Question

I have a matrix and a boolean vector:

>>>from numpy import *
>>>a = arange(20).reshape(4,5)
array([[ 0,  1,  2,  3,  4],
   [ 5,  6,  7,  8,  9],
   [10, 11, 12, 13, 14],
   [15, 16, 17, 18, 19]])

>>>b = asarray( [1, 1, 0, 1] ).reshape(-1,1)
array([[1],
   [1],
   [0],
   [1]])

Now I want to select all the corresponding rows in this matrix where the corresponding index in the vector is equal to zero.

>>>a[b==0]
array([10])

How can I make it so this returns this particular row?

[10, 11, 12, 13, 14]

Answer 1

The shape of b is somewhat strange, but if you can craft it as a nicer index it's a simple selection:

idx = b.reshape(a.shape[0])
print a[idx==0,:]

>>> [[10 11 12 13 14]]

You can read this as, "select all the rows where the index is 0, and for each row selected take all the columns". Your expected answer should really be a list-of-lists since you are asking for all of the rows that match a criteria.