From b422215601e08101b73bc632eff5b7ea37e4ef47 Mon Sep 17 00:00:00 2001 From: "Jessica S. Yu" <15913767+jessicasyu@users.noreply.github.com> Date: Fri, 17 Feb 2023 13:34:22 -0500 Subject: [PATCH] Fix calculate shape stats to use same coeff columns (#9) --- src/abm_shape_collection/calculate_shape_stats.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/src/abm_shape_collection/calculate_shape_stats.py b/src/abm_shape_collection/calculate_shape_stats.py index 90b476e..ecb5266 100644 --- a/src/abm_shape_collection/calculate_shape_stats.py +++ b/src/abm_shape_collection/calculate_shape_stats.py @@ -11,8 +11,9 @@ def calculate_shape_stats( ) -> pd.DataFrame: all_stats = [] - data_transform = pca.transform(data.filter(like="shcoeffs").values) - ref_data_transform = pca.transform(ref_data.filter(like="shcoeffs").values) + columns = ref_data.filter(like="shcoeffs").columns + data_transform = pca.transform(data[columns].values) + ref_data_transform = pca.transform(ref_data[columns].values) for component in range(components): ks_stats = get_ks_statistic(data_transform[:, component], ref_data_transform[:, component]) @@ -20,7 +21,7 @@ def calculate_shape_stats( all_stats.append(ks_stats) for tick, tick_data in data.groupby("TICK"): - tick_data_transform = pca.transform(tick_data.filter(like="shcoeffs").values) + tick_data_transform = pca.transform(tick_data[columns].values) for component in range(components): tick_ks_stats = get_ks_statistic(