Replacing value in python numpy masked array

Question

I have the foll. numpy masked array:

masked_array(data =
 [[-- -- -- ..., -- -- --]
 [-- -- -- ..., -- -- --]
 [-- -- -- ..., -- -- --]
 ..., 
 [-- -- -- ..., -- -- --]
 [-- -- -- ..., -- -- --]
 [-- -- -- ..., -- -- --]],
             mask =
 [[ True  True  True ...,  True  True  True]
 [ True  True  True ...,  True  True  True]
 [ True  True  True ...,  True  True  True]
 ..., 
 [ True  True  True ...,  True  True  True]
 [ True  True  True ...,  True  True  True]
 [ True  True  True ...,  True  True  True]],
       fill_value = -9999.0)

I want to replace all the -9999.0 in this masked array with 0.0, but the foll. does not work:

arr.data[arr == -9999.0] = 0.0

The resulting arr still has all the -9999.0 in it. How do I fix it?

--EDIT:

This is what arr.data looks like:

array([[-9999., -9999., -9999., ..., -9999., -9999., -9999.],
       [-9999., -9999., -9999., ..., -9999., -9999., -9999.],
       [-9999., -9999., -9999., ..., -9999., -9999., -9999.],
       ..., 
       [-9999., -9999., -9999., ..., -9999., -9999., -9999.],
       [-9999., -9999., -9999., ..., -9999., -9999., -9999.],
       [-9999., -9999., -9999., ..., -9999., -9999., -9999.]], dtype=float32)

Are you sure it has -9999.0 in it, and not some other floats very close to -9999.0? — wim
– wim, Commented Jan 19, 2017 at 23:07
Do you also want to change the mask? arr.filled(0) might be what you want. — hpaulj
– hpaulj, Commented Jan 20, 2017 at 0:50

wim · Accepted Answer · 2017-01-19 23:11:44Z

3

The mask is on where the arr.mask has value True.

All those values -9999. values are masked.

If you want it to apply to the masked values aswell, instead of using this:

arr.data[arr == -9999.0] = 0.0

It should be this:

arr.data[arr.data == -9999.0] = 0.0

Note: Be careful with float equality comparisons like this. Usually you want comparison within a tolerance, instead of direct equality. See np.isclose to read more details about that.

answered Jan 19, 2017 at 23:11

wim

368k113 gold badges681 silver badges816 bronze badges

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

Collectives™ on Stack Overflow

Replacing value in python numpy masked array

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related