[Numpy-discussion] Changing MaskedArray.squeeze() to never return masked
allanhaldane at gmail.com
Thu Aug 10 13:00:30 EDT 2017
On 07/18/2017 09:52 AM, Benjamin Root wrote:
> This sort of change seems very similar to the np.diag() change a few years
> ago. Are there lessons we could learn from then that we could apply to here?
> Why would the returned view not be a masked array?
> Ben Root
I am in favor of the proposed change below.
I'd like to merge it, but before that I want to make sure I understand
Are you referring to the proposed change to make diag return a view
instead of a copy? Note that this has not actually happened yet:
Also, I think this case is different because it does not change core
numpy, rather this is to make the MaskedArray module act more
consistently with core numpy. Because of that I think it is much less
problematic than the diag changes.
> On Tue, Jul 18, 2017 at 9:37 AM, Eric Wieser <wieser.eric+numpy at gmail.com>
>> When using ndarray.squeeze, a view is returned, which means you can do
>> the follow (somewhat-contrived) operation:
>>>>> def fill_contrived(a):
>> a.squeeze()[...] = 2
>> return a
>> However, when tried with a masked array, this can fail, breaking liskov
>>>>> fill_contrived(np.ma.array(, mask=[True]))
>> MaskError: Cannot alter the masked element.
>> This fails because squeeze breaks the contract of returning a view,
>> instead deciding sometimes to return masked.
>> There is a patch that fixes this in gh-9432
>> <https://github.com/numpy/numpy/pull/9432> - however, by necessity it
>> breaks any existing code that uses m_arr.squeeze() is np.ma.masked.
>> Is this too breaking a change?
>> NumPy-Discussion mailing list
>> NumPy-Discussion at python.org
> NumPy-Discussion mailing list
> NumPy-Discussion at python.org
More information about the NumPy-Discussion