Question about the evaluation

[luckyyangrun]

Dear organizer:
What is the correct shape of prediction?

Our code (under PyTorch framework) returns a numpy.ndarray with shape [134, 288, 1] through forecast() interface, but we got the error :
'operands could not be broadcast together with shapes (134,) (144,) '

[zhoujingbo]

Hi,

Please refer to our baseline for test prediction https://github.com/PaddlePaddle/PaddleSpatial/tree/main/apps/wpf_baseline_gru

But sorry that we do not implement it in PyTorch, and we cannot answer it.

Thanks a lot for your understanding.

Regards

[ramber1836]

Dear organizer: What is the correct shape of prediction? Our code (under PyTorch framework) returns a numpy.ndarray with shape [134, 288, 1] through forecast() interface, but we got the error : 'operands could not be broadcast together with shapes (134,) (144,) '

I guess you can repack the returned numpy.ndarray with shape [134, 288, 1] to a collection of numpy.ndarrays where each one with the shape [288, 1]. Hope it works.

[ramber1836]

Correct, the shape of the prediction is [134, 288, 1]. Have you tested your code with the released test toy data?

…

On Sat, May 14, 2022 at 4:23 PM luckyyangrun ***@***.***> wrote: Dear organizer: *What is the correct shape of prediction?* [image: image] <https://user-images.githubusercontent.com/67430592/168417323-d63215c5-712c-462b-a00b-6125d0e6f788.png> Our code (under PyTorch framework) returns a numpy.ndarray with shape [134, 288, 1] through forecast() interface, but we got the error : *'operands could not be broadcast together with shapes (134,) (144,) '* — Reply to this email directly, view it on GitHub <#58>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEVHY37NEXBQEICNMLW4QJ3VJ5PHNANCNFSM5V5H3CVA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>