update numpy array where not masked

Question

My question is twofolded

First, lets say I've two numpy arrays, that are partially masked

array_old
[[-- -- -- --]
 [10 11 -- --]
 [12 14 -- --]
 [-- -- 17 --]]

array_update
[[--  5 -- --]
 [-- --  9 --]
 [-- 15  8 13]
 [-- -- 19 16]]

How to get create a new array, where all non-masked values are updated or ammended, like:

array_new
[[--  5 -- --]
 [10 11  9 --]
 [12 15  8 13]
 [-- -- 19 16]]

Secondly, If possible, how to do above in a 3d numpy array?

UPDATE:

For the second part, now I use a for loop, using @freidrichen method as follows:

array = np.ma.masked_equal([[[0, 0, 0, 0], [10, 11, 0, 0], [12, 14, 0, 0], [0, 0, 17, 0]],[[0, 5, 0, 0], [0, 0, 9, 0], [0, 15, 8, 13], [0, 0, 19, 16]],[[0, 0, 0, 0], [5, 0, 0, 13], [8, 14, 0, 0], [0, 0, 17, 0]],[[6, 7, 8, 9], [0, 0, 0, 0], [0, 0, 0, 21], [0, 0, 0, 0]]], 0)

a = array[0,::]
for ix in range(array.shape[0] - 1):
    b = array[ix,::] 
    c = array[ix+1,::]
    b[~c.mask] = c.compressed()
    a[~b.mask] = b.compressed()

Not sure if that's the best solution

@holdenweb you are right, let me update and yes, where new non-masked values are available it should overwrite existing ones — Mattijn
– Mattijn, Commented Oct 5, 2016 at 9:45

freidrichen · Accepted Answer · 2016-10-05 09:12:45Z

5

Use a[~b.mask] = b.compressed().

a[~b.mask] selects all the values in a where b is not masked. b.compressed() is a flattened array with all the non-masked values in b.

Example:

>>> a = np.ma.masked_equal([[0, 0, 0, 0], [10, 11, 0, 0], [12, 14, 0, 0], [0, 0, 17, 0]], 0)
>>> b = np.ma.masked_equal([[0, 5, 0, 0], [0, 0, 9, 0], [0, 15, 8, 13], [0, 0, 19, 16]], 0)
>>> a[~b.mask] = b.compressed()
>>> a
[[-- 5 -- --]
[10 11 9 --]
[12 15 8 13]
[-- -- 19 16]]

This should work with 3d arrays too.

answered Oct 5, 2016 at 9:12

freidrichen

2,6041 gold badge22 silver badges25 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Mattijn Over a year ago

Thanks for your great answer! What I actually ment with 3d arrays, is that one 3d array contains several 2d arrays/slices [as array_update1, array_update2], where the output is a 2d array filled with the most 'recent' values [and maybe no masked values anymore]

Collectives™ on Stack Overflow

update numpy array where not masked

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related