load_metrics_csv_as_df

behavenet.plotting.load_metrics_csv_as_df(hparams, lab, expt, metrics_list, test=False, version='best', version_dir=None)

Load metrics csv file and return as a pandas dataframe for easy plotting.

Parameters:

• hparams (dict) – requires sessions_csv, multisession, lab, expt, animal and session

• lab (str) – for get_lab_example

• expt (str) – for get_lab_example

• metrics_list (list) – names of metrics to pull from csv; do not prepend with ‘tr’, ‘val’, or ‘test’

• test (bool) – True to only return test values (computed once at end of training)

• version (str) – best to find best model in tt expt, None to find model with hyperparams defined in hparams, int to load specific model

Return type:

pandas.DataFrame object