search

cantinilab / scc_reproducibility

1 stars 0 forks source link

How to reproduce the figures #1

Open ZichongWang opened 3 months ago

ZichongWang commented 3 months ago

Hi! Recently I'm trying to reproduce the results of the paper, especially the baselines. But it seems that the output of run_imputation.py and run_integration.py is consists of a few folders named by algorithms (eg. run-scGLUE) which contains some csv files, and three pickle files: full_results.pickle, mean_results.pickle, median_results.pickle. However, to reproduce the final images, I need to run fig_bench.ipynb and fig_cline.ipynb, which need bench_rez.pickle and cline_rez.pickle as input. The structure and content of output results are simular to neither benchrez and clinerez, and not able to run the notebook. What's the correct procedure to reproduce the figures? Could you please give a more detailed instruction? Thanks.

jules-samaran commented 2 months ago

Hello, thanks for opening this issue! I'm busy with revisions at the moment but I'll answer as soon as I can!

ZichongWang commented 2 months ago

Thanks! Looking forward to that!

On Tue, Apr 23, 2024 at 11:31 PM jules-samaran @.***> wrote:

Hello, thanks for opening this issue! I'm busy with revisions at the moment but I'll answer as soon as I can!

— Reply to this email directly, view it on GitHub https://github.com/cantinilab/scc_reproducibility/issues/1#issuecomment-2072717988, or unsubscribe https://github.com/notifications/unsubscribe-auth/BGXCLFJEF3QH5D7IHVTYH7TY6Z5EZAVCNFSM6AAAAABF6MDAY2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANZSG4YTOOJYHA . You are receiving this because you authored the thread.Message ID: @.***>