Doubt about float16 type support on scipy.signal.convolve2d method

quic_sanirudh <[email protected]> Tue, 21 Apr 2020 06:10:14 +0000

Newsgroups	gmane.comp.python.scientific.user
Message-ID	<CH2PR02MB6805382C24D55812197BD327ECD50@CH2PR02MB6805.namprd02.prod.outlook.com>

Hi,

I'm new to this community and in general to scipy. I tried to use the scipy.signal.convolve2d method with the call shown below, where a and b are of types float16.

scipy.signal.convolve2d(a, b, mode='valid')

I get the below error, and wanted to check if float16 types are not supported in the convolve2d method by design or if this could be a missed feature.

ValueError: convolve2d not available for this type.

Let me know if I have made any mistakes or if this I can perform float16 convolve2d  using a different function perhaps.

Thanks,
Anirudh

_______________________________________________
SciPy-User mailing list
[email protected]
https://mail.python.org/mailman/listinfo/scipy-user