search

USGS-R / river-dl

Deep learning model for predicting environmental variables on river systems
Creative Commons Zero v1.0 Universal
21 stars 15 forks source link

Handle validation/testing sites as well as times in preprocessing #160

Closed jsadler2 closed 2 years ago

jsadler2 commented 2 years ago

Currently, the preprocessing script function prep_all_data handles the partitioning into test/validation/testing sets temporally, as in you give it start and end dates for each partition and it divides the data accordingly.

It would be useful if you could also provide a list of validation and test segments (or sites) so that those sites end up in the validation or test data and are not seen in the training set.

SimonTopp commented 2 years ago

Just a quick +1 to this! Would be very helpful. It might also actually streamline the script a little because it could replace some of the functionality of the exclude file/args.