CODEWITHSUNDEEP
numpynumpy-indexing
user2309803
user2309803

Reputation: 645

NumPy : How to determine the index of the first axis of an ndarray according to some condition?

Consider the following ndarray a -

In [117]: a                                                                                          
Out[117]: 
array([[[nan, nan],
        [nan, nan],
        [nan, nan]],

       [[ 3., 11.],
        [ 7., 13.],
        [12., 16.]],

       [[ 0.,  4.],
        [ 6.,  1.],
        [ 5.,  8.]],

       [[17., 10.],
        [15.,  9.],
        [ 2., 14.]]])

The minimum computed on the first axis is -

In [118]: np.nanmin(a, 0)                                                                            
Out[118]: 
array([[0., 4.],
       [6., 1.],
       [2., 8.]])

which is a[2] from visual inspection. What is the most efficient way to calculate this index 2

Upvotes: 0

Views: 42

Answers (1)

Aly Hosny
Aly Hosny

Reputation: 827

as suggested by @Divakar you can use np.nanargmin

import numpy as np

a = np.array([[[np.nan, np.nan],
        [np.nan, np.nan],
        [np.nan, np.nan]],

       [[ 3., 11.],
        [ 7., 13.],
        [12., 16.]],

       [[ 0.,  4.],
        [ 6.,  1.],
        [ 5.,  8.]],

       [[17., 10.],
        [15.,  9.],
        [ 2., 14.]]])
minIdx = np.nanargmin(np.sum(a,(1,2)))
minIdx
2
a[minIdx]
array([[0., 4.],
       [6., 1.],
       [5., 8.]])

Upvotes: 1

Related Questions