def summarize_with_groups(dframe, stats, group):
"""
Calculate summary statistics for group.
"""
grouped_stats = series_to_jsondict(
dframe.groupby(group).apply(summarize_df))
stats.update({group: grouped_stats})
return stats
def summarize_df(dframe):
"""
Calculate summary statistics
"""
dtypes = dframe.dtypes
# TODO handle empty data
# before we wrapped with filter(lambda d: d[DATA] and len(d[DATA]), array)
return dict([
(col, {
SUMMARY: series_to_jsondict(summarize_series(dtypes[col], data))
}) for col, data in dframe.iteritems() if not col in MONGO_RESERVED_KEYS
])