[FIXED] Boolean indexing on multidimensionnal array

Issue

I'm very new in Python and in Numpy. In fact, I'm just learning.

I'm reading this tutorial, and I got stuck in these lines :

>>> x = np.arange(30).reshape(2,3,5)
>>> x
array([[[ 0,  1,  2,  3,  4],
        [ 5,  6,  7,  8,  9],
        [10, 11, 12, 13, 14]],
       [[15, 16, 17, 18, 19],
        [20, 21, 22, 23, 24],
        [25, 26, 27, 28, 29]]])
>>> b = np.array([[True, True, False], [False, True, True]])
>>> x[b]
array([[ 0,  1,  2,  3,  4],
       [ 5,  6,  7,  8,  9],
       [20, 21, 22, 23, 24],
       [25, 26, 27, 28, 29]])

I can't understand how we have come up with the result of x[b].

I also try to guess the result of x[[False, False, False, True]]

Please explain to me, I'm a very newbie.

Solution

You have 3 arrays in 1 array:

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

With your following line: b = np.array([[True, True, False], ...]) you say that you want to keep the first 2 rows (the first 2 True values) and that you don't want the last row (the last False value).

The other part works the same way, you have 3 arrays in 1 array:

[
 [15, 16, 17, 18, 19],
 [20, 21, 22, 23, 24],
 [25, 26, 27, 28, 29]
]

And your line b = np.array([..., [False, True, True]]) says to not keep the first row (because first value is False) but that you want to keep the two last lines (2 last values are True).

Answered By - Stanko