Batch runs

[aspeake]

Fixes https://github.com/trynthink/scout/issues/367

Enable running multiple ymls with a single script. This PR adds a new package module run_batch.py that accepts a directory of yml configuration files and runs ecm_prep.py and run.py.

To run: python scout/run_batch.py -b <my_yml_dir>

Summary of updates:

• As a performance optimization, run_batch.py groups together ymls with ecm_prep arguments that are equal with the exception of ECM downselect arguments (ecm_files, ecm_directory, and ecm_files_regex), and runs ecm_prep.py just once for each group.
• After running ecm_prep for a group, run.py is run for each yml in that group
• New optional --results_directory argument for run.py to specify where results & plots are output
• Each yml gets its own results directory, if not provided in --results_directory, it will default to ./results/<name_of_yml>/
• Each "yml group" gets a custom directory within ./generated for intermediate files

[aspeake]

Depending on timing of https://github.com/trynthink/scout/pull/375, it may be good to update fp.LOGS path based on which yml is being run: https://github.com/trynthink/scout/pull/375#discussion_r1543487672

[aspeake]

@trynthink do you think its important to be more clear about which files in ./generated belong to which yml/yml group? It is not at all apparent right now, but are those files ever important to users for debugging?

[aspeake]

@trynthink do you think its important to be more clear about which files in ./generated belong to which yml/yml group? It is not at all apparent right now, but are those files ever important to users for debugging?

Copy ymls into their ./generated directories and rename directories as "batch_run1", "batch_run2"...?