Current implementation of extract_results relies on get_scenario_info to extract information about the scenarios; the latter function only stores the number of draws and the number of runs, which is retrieved from the first draw folder.
This can be problematic if we have results from a partially completed batch-run e.g. if I have submitted 10 jobs, and draw 5 has finished first, I'd like to be able to use extract_results on these results.
I think we could do this if we allow users to pass the draw and run as an argument to extract_results. It might also be useful for get_scenario_info to not store the number of draws but their actual names, so that we can read what we have.
Current implementation of
extract_resultsrelies onget_scenario_infoto extract information about the scenarios; the latter function only stores the number of draws and the number of runs, which is retrieved from the first draw folder.This can be problematic if we have results from a partially completed batch-run e.g. if I have submitted 10 jobs, and draw 5 has finished first, I'd like to be able to use
extract_resultson these results.I think we could do this if we allow users to pass the draw and run as an argument to
extract_results. It might also be useful forget_scenario_infoto not store the number of draws but their actual names, so that we can read what we have.