add some of the helper functions from experiment scripts into package e

[abigailsnyder]

eg

# helper function to pull the GSAT data of the specific ensemble members in
# a target data frame:
def get_orig_GSAT_data(target_df): 
    ...

# Function to remove any ensemble members from a target data frame that
# stop before 2099, for example, ending in 2014 like some MIROC6 SSP245:
def prep_target_data(target_df):
    ...

[abigailsnyder]

turn code block for pulling off first 5 (N) numerical ensemble members from a target set into a package function:

# we will target  with the first 5 numerical ensemble members.
    # Not all models start the ensemble count at 1,
    # And not all experiments of a given model report the
    # same realizations.
    # select 5 ensemble realizations to
    # look at if there are more than 5.
    f = lambda x: x.ensemble[:x.idx]

    if not target_data.empty:
        ensemble_list = pd.DataFrame({'ensemble': target_data["ensemble"].unique()})
        ensemble_list['idx'] = ensemble_list['ensemble'].str.index('i')
        ensemble_list['ensemble_id'] = ensemble_list.apply(f, axis=1)
        ensemble_list['ensemble_id'] = ensemble_list['ensemble_id'].str[1:].astype(int)
        ensemble_list = ensemble_list.sort_values('ensemble_id').copy()
        if len(ensemble_list) > 5:
            ensemble_keep = ensemble_list.iloc[0:5].ensemble
        else:
            ensemble_keep = ensemble_list.ensemble

        target_data = target_data[target_data['ensemble'].isin(ensemble_keep)].copy()
        del (ensemble_keep)
        del (ensemble_list)

[abigailsnyder]

filtering one data frame based on another data frame

# Keep only the entries that appeared in pangeo_good_ensembles:
keys =['model', 'experiment', 'ensemble']
i1 = data.set_index(keys).index
i2 = pangeo_good_ensembles.set_index(keys).index
data= data[i1.isin(i2)].copy()
del(i1)
del(i2)