numpy boolean indexing multiple conditions

Question

I have a two dimensional numpy array and I am using python 3.5. I am starting to learn about Boolean indexing which is way cool. I can do this with my two dimensional array, arr. mask = arr > 127 arr[mask] = 0

This works perfect but now I am trying to change this logic to use boolean indexing

for x in range(arr.shape[0]):
    for y in range(arr.shape[1]):
        if arr[x,y] < -10:
            arr[x,y] = 0
        elif arr[x,y] < 15:
            arr[x,y] = arr[x,y] + 5
        else:
            arr[x,y] = 30

I tried multiple conditional operators for my indexing but I get the following error:

ValueError: boolean index array should have 1 dimension boolean index array should have 1 dimension.

I tried multiple versions to try to get this to work. Here is one try that produced the ValueError.

 arr_temp = arr.copy()
 mask = arry_temp < -10
 mask2 = arry_temp < 15
 mask3 = mask ^ mask3
 arr[mask] = 0
 arr[mask3] = arry[mask3] + 5
 arry[~mask2] = 30 

I received the error on mask3. I am new to this so I know the code above is not efficient trying to work out it.

Any tips would be appreciated.

Answer 1

This might help. Consider a numpy array of floating point values foo.

import numpy as np
foo=np.array([0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8])

foo yields

array([0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8, 0.8])

This is how you get the values in foo > 0.3

foo[np.where( foo > 0.3)]

yields

array([0.4, 0.5, 0.6, 0.7, 0.8])

This is how to do the same with multiple conditions. In this case, values > 0.3 and less than 0.6.

foo[np.logical_and(foo > 0.3, foo < 0.6)]

yields

array([0.4, 0.5])

Alternatively using boolean mask array

mask_1 = foo > 0.3
mask_2 = foo < 0.6
mask_3 = np.logical_and(mask_1, mask_2)
mask_3

Yields a boolean mask array

array([False, False, False,  True,  True, False])

Which you can then use to slice the array via

foo[mask_3]

Yields

array([0.4, 0.5])