You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In reviewing #711 it was identified that there is no use for the returned first_marked_mask as it is not used in any subsequent processing. Commit #1ffe0c therefore assigns it to _ which is discarded.
Rather than return something and discard it we should perhaps consider removing the mask from being returned at all and changing line 434 to...
returnimg
As an aside I noticed the typehint for the remove_scars() function doesn't have an explicit return type which if this change is made would be npt.NDArray (I discovered Numpy now has a specific numpy.typing module which if import numpy.typing as npt would allow use of the generic NDArray). The docstring also says self.img is returned when it should, after this change, just be...
Returns
-------
npt.NDArray
...
You're thoughts on whether there is ever any use for the scar mask would be appreciated @SylviaWhittle.
The text was updated successfully, but these errors were encountered:
So I think we do use the returned scar mask in the second instance, where one of the images for the Filters' images dictionary is set to "scar_mask":
But it isn't plotted even with 'all' as the image set.
When developing this feature I had thought it would be of interest to users to see the scars that the scar removal had removed, however given that it was never plotted and that nobody has asked / complained about it, perhaps it should be removed?
I personally have used this a couple times, but not more than that, and as a developer, it's simple to add it back in when needed.
What do you think about keeping these arrays or not? I noticed mention in #675 about processing .asd taking up ~2GB of RAM so keeping these would only inflate that.
In reviewing #711 it was identified that there is no use for the returned
first_marked_mask
as it is not used in any subsequent processing. Commit #1ffe0c therefore assigns it to_
which is discarded.Rather than return something and discard it we should perhaps consider removing the mask from being returned at all and changing line 434 to...
As an aside I noticed the typehint for the
remove_scars()
function doesn't have an explicit return type which if this change is made would benpt.NDArray
(I discovered Numpy now has a specific numpy.typing module which ifimport numpy.typing as npt
would allow use of the genericNDArray
). The docstring also saysself.img
is returned when it should, after this change, just be...You're thoughts on whether there is ever any use for the scar mask would be appreciated @SylviaWhittle.
The text was updated successfully, but these errors were encountered: