I recommend a change to how the ouput filename is constructed in the do_seg() workflow. The current method for creating the output filename is as follows:
If you pass an alternate output directory to do_seg stored in the sample_direc variable, the outputs are still saved to the directory where the input files are stored. I would like to be able to pass an alternate directory to save model predictions. Therefore, I recommend changing:
This is a good idea, and thanks for suggesting it, but I decline to make this change because it will break downstream applications. User can move folders if they need to.
I recommend a change to how the ouput filename is constructed in the
do_seg()
workflow. The current method for creating the output filename is as follows:https://github.com/Doodleverse/doodleverse_utils/blob/83128ee3e7a8dd901415c6f00145f31b8c44bb02/doodleverse_utils/prediction_imports.py#L128-L154
If you pass an alternate output directory to
do_seg
stored in thesample_direc
variable, the outputs are still saved to the directory where the input files are stored. I would like to be able to pass an alternate directory to save model predictions. Therefore, I recommend changing:https://github.com/Doodleverse/doodleverse_utils/blob/83128ee3e7a8dd901415c6f00145f31b8c44bb02/doodleverse_utils/prediction_imports.py#L151-L154
to:
Just a recommendation though!!