`highpass_spatial_filter` returns as `float32`

[JoeZiminski]

I just realised highpass_spatial_filter is returning as float32. here. (cast during scipy filtering I think). Should this be cast to self.dtype similar to filter.py?

I also see the uint is forced to float32 in filtering here. Should this also be a consideration for highpass_spatial_filter?

[h-mayorquin]

Did you ever make your mind about this? Any suggestion on what would be better?

[JoeZiminski]

I wasn't sure about the best approach, although my guess is that copying the approach in filter.py where it is cast back to self.dtype on return is best?

[alejoe91]

@h-mayorquin @JoeZiminski this is fixed in main! Getting an int16 out of an int16 input :)

[alejoe91]

@h-mayorquin @JoeZiminski this is fixed in main! Getting an int16 out of an int16 input :)

Forget about it! the traces are indeed in float. Will make a PR with a fix!

[JoeZiminski]

That's great thanks @alejoe91 and thanks for reviving this @h-mayorquin!

[h-mayorquin]

Great. Should we close this?

[alejoe91]

I'll make a PR today to fix it