Closed peicai closed 7 months ago
I fear this will be a mess to merge. Can you remove tracking on all files outside stLearn? - is there a reason for deleting SpaGCN configs?
Regarding SpaGCN configs, I might have deleted them to resolve conflicts somehow. I've updated all files in the method from the main branch, hoping that it will make merging easier.
Can you remove tracking on all files outside stLearn I removed tracking on all other files, but I'm not sure if I did it correctly or if it will work.
Oh no, that made it much worse - you need to remove tracking of all files outside stLearn, not add tracking.
Something like
git reset --soft <commit>
to a specific commit without external changes and then possibly remove staged files outside stLearn with:
git restore --staged <files>
The goal is to reduce Files changed to only have stLearn files. Otherwise, merging will be a mess. Hope it makes sense :)
Thanks for your clear explanation and codes. Unfortunately, I made external changes in the first few commits. It would be much easier to close and recreate a branch for stLearn and resend the pull request.
Hi @peicai, I would close this PR and when u have time could you branch the native branch and update the implementation of stLearn in the format of native implementation (I think you basically get it done) and submit a PR to the native branch? Thanks!
Added a script for stLearn implementation, a yml and 13 config files; linked to issue https://github.com/SpatialHackathon/SpaceHack2023/issues/6.
Note: 11 out of 13 config files require the inclusion of image paths, specifically for 'H_E.tiff' and 'H_E.json'. The proper formatting example for these image paths is as follows: --image $data_dir/H_E.tiff $data_dir/H_E.json. In Snakemake workflows, the function
get_sample_image
in themethods.smk
file might need to be modified.