CODEWITHSUNDEEP
pythonnumpymultidimensional-array
Reactormonk
Reactormonk

Reputation: 21690

Subset of an ndarray based on another array

I have an array of ints, they need to be grouped by 4 each. I'd also like to select them based on another criterion, start < t < stop. I tried

data[group].reshape((-1,4))[start < t < stop]

but that complains about the start < t < stop because that's hardcoded syntax. Can I somehow intersect the two arrays from start < t and t < stop?

Upvotes: 2

Views: 104

Answers (2)

CT Zhu
CT Zhu

Reputation: 54330

The right way of boolean indexing for an array should be like this:

>>> import numpy as np
>>> a=np.random.randint(0,20,size=24)
>>> b=np.arange(24)
>>> b[(8<a)&(a<15)] #rather than 8<a<15
array([ 3,  5,  6, 11, 13, 16, 17, 18, 20, 21, 22, 23])

But you may not be able to reshape the resulting array into a shape of (-1,4), it is a coincidence that the resulting array here contains 3*4 elements.

EDIT, now I understand your OP better. You always reshape data[group] first, right?:

>>> b=np.arange(96)
>>> b.reshape((-1,4))[(8<a)&(a<15)]
array([[12, 13, 14, 15],
       [20, 21, 22, 23],
       [24, 25, 26, 27],
       [44, 45, 46, 47],
       [52, 53, 54, 55],
       [64, 65, 66, 67],
       [68, 69, 70, 71],
       [72, 73, 74, 75],
       [80, 81, 82, 83],
       [84, 85, 86, 87],
       [88, 89, 90, 91],
       [92, 93, 94, 95]])

Upvotes: 2

John Zwinck
John Zwinck

Reputation: 249123

How about this?

import numpy as np

arr = np.arange(32)
t = np.arange(300, 364, 2)
start = 310
stop = 352
mask = np.logical_and(start < t, t < stop)
print mask
print arr[mask].reshape((-1,4))

I did the masking before the reshaping, not sure if that's what you wanted. The key part is probably the logical_and().

Upvotes: 2

Related Questions