search

USGS-R / river-dl

Deep learning model for predicting environmental variables on river systems
Creative Commons Zero v1.0 Universal
21 stars 14 forks source link

More flexible combined metrics #189

Closed jsadler2 closed 2 years ago

jsadler2 commented 2 years ago

This PR makes two main changes in the combined_metrics function.

  1. it allows the user to provide any list of arbitrary ids to distinguish metrics files
  2. it allows the user to pass any number of prediction datasets with any labels to combine_metrics instead of only pred_trn, pred_val, and pred_tst

closes #183 closes #184

SimonTopp commented 2 years ago

This looks good to me Jeff! Are you thinking that the unique IDs could be generated as wildcards in the snakefile or something? Or otherwise in some standardized format? I wonder if it'd be worthwhile to set some default where if no ID is passed it's automatically set to the current date or git commit hash or something.

jsadler2 commented 2 years ago

Thanks for taking a look, @SimonTopp!

Yes. I was thinking that the ID cols could be populated with snakemake wildcards.

I'm hesitant to supply a default since it may not be wanted or needed by every project. I think it's okay to leave that to the project if they want to supply the date or git commit hash.