Source code for simba.feature_extractors.feature_extractor_14bp
__author__ = "Simon Nilsson; sronilsson@gmail.com"
import math
import os
from collections import defaultdict
from copy import deepcopy
import numpy as np
import pandas as pd
from simba.feature_extractors.perimeter_jit import jitted_hull
from simba.mixins.config_reader import ConfigReader
from simba.mixins.feature_extraction_mixin import FeatureExtractionMixin
from simba.utils.enums import Formats
from simba.utils.printing import SimbaTimer, stdout_success
from simba.utils.read_write import get_fn_ext, read_df, write_df
[docs]class ExtractFeaturesFrom14bps(ConfigReader, FeatureExtractionMixin):
"""
Extracts hard-coded set of features from pose-estimation data from two animals with 7 tracked body-parts each.
Results are stored in the `project_folder/csv/features_extracted` directory of the SimBA project.
.. note::
* `Feature extraction tutorial <https://github.com/sgoldenlab/simba/blob/master/docs/tutorial.md#step-5-extract-features>`__.
* `Expected pose configuration <https://github.com/sgoldenlab/simba/blob/master/simba/pose_configurations/schematics/6.png>`_
.. image:: _static/img/pose_configurations/6.png
:alt: 6
:width: 300
:align: center
:param str config_path: path to SimBA project config file in Configparser format
:example:
>>> feature_extractor = ExtractFeaturesFrom14bps(config_path='MyProjectConfig')
>>> feature_extractor.run()
"""
def __init__(self, config_path: str):
FeatureExtractionMixin.__init__(self, config_path=config_path)
ConfigReader.__init__(self, config_path=config_path)
self.in_headers = self.get_feature_extraction_headers(
pose="2 animals 14 body-parts"
)
self.mouse_1_headers, self.mouse_2_headers = (
self.in_headers[0:21],
self.in_headers[21:],
)
self.mouse_2_p_headers = [x for x in self.mouse_2_headers if x[-2:] == "_p"]
self.mouse_1_p_headers = [x for x in self.mouse_1_headers if x[-2:] == "_p"]
self.mouse_1_headers = [x for x in self.mouse_1_headers if x[-2:] != "_p"]
self.mouse_2_headers = [x for x in self.mouse_2_headers if x[-2:] != "_p"]
print(
"Extracting features from {} file(s)...".format(str(len(self.files_found)))
)
[docs] def run(self):
"""
Method to compute and save features to disk. Results are saved in the `project_folder/csv/features_extracted`
directory of the SimBA project.
Returns
-------
None
"""
for file_cnt, file_path in enumerate(self.files_found):
video_timer = SimbaTimer(start=True)
roll_windows = []
_, self.video_name, _ = get_fn_ext(file_path)
video_settings, self.px_per_mm, fps = self.read_video_info(
video_name=self.video_name
)
for window in self.roll_windows_values:
roll_windows.append(int(fps / window))
self.in_data = (
read_df(file_path, self.file_type)
.fillna(0)
.apply(pd.to_numeric)
.reset_index(drop=True)
)
print(
"Processing {} ({} frames)...".format(
self.video_name, str(len(self.in_data))
)
)
self.in_data = self.insert_default_headers_for_feature_extraction(
df=self.in_data,
headers=self.in_headers,
pose_config="14 body-parts",
filename=file_path,
)
self.out_data = deepcopy(self.in_data)
mouse_1_ar = np.reshape(
self.out_data[self.mouse_1_headers].values,
(len(self.out_data / 2), -1, 2),
).astype(np.float32)
self.out_data["Mouse_1_poly_area"] = (
jitted_hull(points=mouse_1_ar, target=Formats.PERIMETER.value)
/ self.px_per_mm
)
mouse_2_ar = np.reshape(
self.out_data[self.mouse_2_headers].values,
(len(self.out_data / 2), -1, 2),
).astype(np.float32)
self.out_data["Mouse_2_poly_area"] = (
jitted_hull(points=mouse_2_ar, target=Formats.PERIMETER.value)
/ self.px_per_mm
)
self.in_data_shifted = (
self.out_data.shift(periods=1).add_suffix("_shifted").fillna(0)
)
self.in_data = (
pd.concat([self.in_data, self.in_data_shifted], axis=1, join="inner")
.fillna(0)
.reset_index(drop=True)
)
self.out_data["Mouse_1_nose_to_tail"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Tail_base_1_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Tail_base_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_nose_to_tail"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Tail_base_2_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Tail_base_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_width"] = self.euclidean_distance(
self.out_data["Lat_left_1_x"].values,
self.out_data["Lat_right_1_x"].values,
self.out_data["Lat_left_1_y"].values,
self.out_data["Lat_right_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_width"] = self.euclidean_distance(
self.out_data["Lat_left_2_x"].values,
self.out_data["Lat_right_2_x"].values,
self.out_data["Lat_left_2_y"].values,
self.out_data["Lat_right_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_Ear_distance"] = self.euclidean_distance(
self.out_data["Ear_left_1_x"].values,
self.out_data["Ear_right_1_x"].values,
self.out_data["Ear_left_1_y"].values,
self.out_data["Ear_right_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_Ear_distance"] = self.euclidean_distance(
self.out_data["Ear_left_2_x"].values,
self.out_data["Ear_right_2_x"].values,
self.out_data["Ear_left_2_y"].values,
self.out_data["Ear_right_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_Nose_to_centroid"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Center_1_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Center_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_Nose_to_centroid"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Center_2_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Center_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_Nose_to_lateral_left"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Lat_left_1_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Lat_left_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_Nose_to_lateral_left"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Lat_left_2_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Lat_left_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_Nose_to_lateral_right"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Lat_right_1_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Lat_right_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_Nose_to_lateral_right"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Lat_right_2_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Lat_right_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_Centroid_to_lateral_left"] = self.euclidean_distance(
self.out_data["Center_1_x"].values,
self.out_data["Lat_left_1_x"].values,
self.out_data["Center_1_y"].values,
self.out_data["Lat_left_1_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_2_Centroid_to_lateral_left"] = self.euclidean_distance(
self.out_data["Center_2_x"].values,
self.out_data["Lat_left_2_x"].values,
self.out_data["Center_2_y"].values,
self.out_data["Lat_left_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_Centroid_to_lateral_right"] = (
self.euclidean_distance(
self.out_data["Center_1_x"].values,
self.out_data["Lat_right_1_x"].values,
self.out_data["Center_1_y"].values,
self.out_data["Lat_right_1_y"].values,
self.px_per_mm,
)
)
self.out_data["Mouse_2_Centroid_to_lateral_right"] = (
self.euclidean_distance(
self.out_data["Center_2_x"].values,
self.out_data["Lat_right_2_x"].values,
self.out_data["Center_2_y"].values,
self.out_data["Lat_right_2_y"].values,
self.px_per_mm,
)
)
self.out_data["Centroid_distance"] = self.euclidean_distance(
self.out_data["Center_2_x"].values,
self.out_data["Center_1_x"].values,
self.out_data["Center_2_y"].values,
self.out_data["Center_1_y"].values,
self.px_per_mm,
)
self.out_data["Nose_to_nose_distance"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Nose_1_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Nose_1_y"].values,
self.px_per_mm,
)
self.out_data["M1_Nose_to_M2_lat_left"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Lat_left_2_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Lat_left_2_y"].values,
self.px_per_mm,
)
self.out_data["M1_Nose_to_M2_lat_left"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Lat_left_2_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Lat_left_2_y"].values,
self.px_per_mm,
)
self.out_data["M1_Nose_to_M2_lat_right"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Lat_right_2_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Lat_right_2_y"].values,
self.px_per_mm,
)
self.out_data["M2_Nose_to_M1_lat_left"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Lat_left_1_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Lat_left_1_y"].values,
self.px_per_mm,
)
self.out_data["M2_Nose_to_M1_lat_right"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Lat_right_1_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Lat_right_1_y"].values,
self.px_per_mm,
)
self.out_data["M1_Nose_to_M2_tail_base"] = self.euclidean_distance(
self.out_data["Nose_1_x"].values,
self.out_data["Tail_base_2_x"].values,
self.out_data["Nose_1_y"].values,
self.out_data["Tail_base_2_y"].values,
self.px_per_mm,
)
self.out_data["M2_Nose_to_M1_tail_base"] = self.euclidean_distance(
self.out_data["Nose_2_x"].values,
self.out_data["Tail_base_1_x"].values,
self.out_data["Nose_2_y"].values,
self.out_data["Tail_base_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_centroid"] = self.euclidean_distance(
self.in_data["Center_1_x_shifted"].values,
self.in_data["Center_1_x"].values,
self.in_data["Center_1_y_shifted"].values,
self.in_data["Center_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_centroid"] = self.euclidean_distance(
self.in_data["Center_2_x_shifted"].values,
self.in_data["Center_2_x"].values,
self.in_data["Center_2_y_shifted"].values,
self.in_data["Center_2_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_nose"] = self.euclidean_distance(
self.in_data["Nose_1_x_shifted"].values,
self.in_data["Nose_1_x"].values,
self.in_data["Nose_1_y_shifted"].values,
self.in_data["Nose_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_nose"] = self.euclidean_distance(
self.in_data["Nose_2_x_shifted"].values,
self.in_data["Nose_2_x"].values,
self.in_data["Nose_2_y_shifted"].values,
self.in_data["Nose_2_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_tail_base"] = self.euclidean_distance(
self.in_data["Tail_base_1_x_shifted"].values,
self.in_data["Tail_base_1_x"].values,
self.in_data["Tail_base_1_y_shifted"].values,
self.in_data["Tail_base_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_tail_base"] = self.euclidean_distance(
self.in_data["Tail_base_2_x_shifted"].values,
self.in_data["Tail_base_2_x"].values,
self.in_data["Tail_base_2_y_shifted"].values,
self.in_data["Tail_base_2_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_left_ear"] = self.euclidean_distance(
self.in_data["Ear_left_1_x_shifted"].values,
self.in_data["Ear_left_1_x"].values,
self.in_data["Ear_left_1_y_shifted"].values,
self.in_data["Ear_left_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_left_ear"] = self.euclidean_distance(
self.in_data["Ear_left_2_x_shifted"].values,
self.in_data["Ear_left_2_x"].values,
self.in_data["Ear_left_2_y_shifted"].values,
self.in_data["Ear_left_2_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_right_ear"] = self.euclidean_distance(
self.in_data["Ear_right_1_x_shifted"].values,
self.in_data["Ear_right_1_x"].values,
self.in_data["Ear_right_1_y_shifted"].values,
self.in_data["Ear_right_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_right_ear"] = self.euclidean_distance(
self.in_data["Ear_right_2_x_shifted"].values,
self.in_data["Ear_right_2_x"].values,
self.in_data["Ear_right_2_y_shifted"].values,
self.in_data["Ear_right_2_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_lateral_left"] = self.euclidean_distance(
self.in_data["Lat_left_1_x_shifted"].values,
self.in_data["Lat_left_1_x"].values,
self.in_data["Lat_left_1_y_shifted"].values,
self.in_data["Lat_left_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_lateral_left"] = self.euclidean_distance(
self.in_data["Lat_left_2_x_shifted"].values,
self.in_data["Lat_left_2_x"].values,
self.in_data["Lat_left_2_y_shifted"].values,
self.in_data["Lat_left_2_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_1_lateral_right"] = self.euclidean_distance(
self.in_data["Lat_right_1_x_shifted"].values,
self.in_data["Lat_right_1_x"].values,
self.in_data["Lat_right_1_y_shifted"].values,
self.in_data["Lat_right_1_y"].values,
self.px_per_mm,
)
self.out_data["Movement_mouse_2_lateral_right"] = self.euclidean_distance(
self.in_data["Lat_right_2_x_shifted"].values,
self.in_data["Lat_right_2_x"].values,
self.in_data["Lat_right_2_y_shifted"].values,
self.in_data["Lat_right_2_y"].values,
self.px_per_mm,
)
self.out_data["Mouse_1_polygon_size_change"] = (
self.in_data["Mouse_1_poly_area_shifted"]
- self.out_data["Mouse_1_poly_area"]
)
self.out_data["Mouse_2_polygon_size_change"] = (
self.in_data["Mouse_2_poly_area_shifted"]
- self.out_data["Mouse_2_poly_area"]
)
print("Calculating hull variables...")
mouse_1_array, mouse_2_array = (
self.in_data[self.mouse_1_headers].to_numpy(),
self.in_data[self.mouse_2_headers].to_numpy(),
)
self.hull_dict = defaultdict(list)
for cnt, (animal_1, animal_2) in enumerate(
zip(mouse_1_array, mouse_2_array)
):
animal_1, animal_2 = np.reshape(animal_1, (-1, 2)), np.reshape(
animal_2, (-1, 2)
)
animal_1_dist, animal_2_dist = self.cdist(
animal_1, animal_1
), self.cdist(animal_2, animal_2)
animal_1_dist, animal_2_dist = (
animal_1_dist[animal_1_dist != 0],
animal_2_dist[animal_2_dist != 0],
)
for animal, animal_name in zip(
[animal_1_dist, animal_2_dist], ["M1", "M2"]
):
self.hull_dict[
"{}_hull_large_euclidean".format(animal_name)
].append(np.amax(animal, initial=0) / self.px_per_mm)
self.hull_dict[
"{}_hull_small_euclidean".format(animal_name)
].append(
np.min(
animal,
initial=self.hull_dict[
"{}_hull_large_euclidean".format(animal_name)
][-1],
)
/ self.px_per_mm
)
self.hull_dict["{}_hull_mean_euclidean".format(animal_name)].append(
np.mean(animal) / self.px_per_mm
)
self.hull_dict["{}_hull_sum_euclidean".format(animal_name)].append(
np.sum(animal) / self.px_per_mm
)
for k, v in self.hull_dict.items():
self.out_data[k] = v
self.out_data["Sum_euclidean_distance_hull_M1_M2"] = (
self.out_data["M1_hull_sum_euclidean"]
+ self.out_data["M2_hull_sum_euclidean"]
)
self.out_data["Total_movement_centroids"] = (
self.out_data["Movement_mouse_1_centroid"]
+ self.out_data["Movement_mouse_2_centroid"]
)
mouse_1_movement_cols = [
x for x in self.out_data.columns if x.startswith("Movement_mouse_1_")
]
mouse_2_movement_cols = [
x for x in self.out_data.columns if x.startswith("Movement_mouse_2_")
]
self.out_data["Total_movement_all_bodyparts_M1"] = self.out_data[
mouse_1_movement_cols
].sum(axis=1)
self.out_data["Total_movement_all_bodyparts_M2"] = self.out_data[
mouse_2_movement_cols
].sum(axis=1)
self.out_data["Total_movement_all_bodyparts_both_mice"] = (
self.out_data["Total_movement_all_bodyparts_M1"]
+ self.out_data["Total_movement_all_bodyparts_M2"]
)
print("Calculating rolling windows: medians, medians, and sums...")
for window in self.roll_windows_values:
col_name = "Sum_euclid_distances_hull_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Sum_euclidean_distance_hull_M1_M2"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Sum_euclid_distances_hull_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Sum_euclidean_distance_hull_M1_M2"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Sum_euclid_distances_hull_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Sum_euclidean_distance_hull_M1_M2"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Movement_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_movement_centroids"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Movement_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_movement_centroids"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Movement_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_movement_centroids"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Distance_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Centroid_distance"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Distance_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Centroid_distance"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Distance_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Centroid_distance"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse1_width_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Mouse_1_width"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse1_width_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Mouse_1_width"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse1_width_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Mouse_1_width"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse2_width_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Mouse_2_width"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse2_width_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Mouse_2_width"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse2_width_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Mouse_2_width"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse1_mean_euclid_distances_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M1_hull_mean_euclidean"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse1_mean_euclid_distances_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M1_hull_mean_euclidean"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse1_mean_euclid_distances_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M1_hull_mean_euclidean"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse2_mean_euclid_distances_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M2_hull_mean_euclidean"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse2_mean_euclid_distances_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M2_hull_mean_euclidean"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse2_mean_euclid_distances_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M2_hull_mean_euclidean"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse1_smallest_euclid_distances_median_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["M1_hull_small_euclidean"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse1_smallest_euclid_distances_mean_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["M1_hull_small_euclidean"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse1_smallest_euclid_distances_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M1_hull_small_euclidean"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse2_smallest_euclid_distances_median_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["M2_hull_small_euclidean"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse2_smallest_euclid_distances_mean_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["M2_hull_small_euclidean"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse2_smallest_euclid_distances_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M2_hull_small_euclidean"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse1_largest_euclid_distances_median_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["M1_hull_large_euclidean"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse1_largest_euclid_distances_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M1_hull_large_euclidean"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse1_largest_euclid_distances_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M1_hull_large_euclidean"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Mouse2_largest_euclid_distances_median_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["M2_hull_large_euclidean"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Mouse2_largest_euclid_distances_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M2_hull_large_euclidean"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Mouse2_largest_euclid_distances_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["M2_hull_large_euclidean"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Total_movement_all_bodyparts_both_mice_median_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["Total_movement_all_bodyparts_both_mice"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Total_movement_all_bodyparts_both_mice_mean_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["Total_movement_all_bodyparts_both_mice"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Total_movement_all_bodyparts_both_mice_sum_{}".format(
str(window)
)
self.out_data[col_name] = (
self.out_data["Total_movement_all_bodyparts_both_mice"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Total_movement_centroids_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_movement_centroids"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Total_movement_centroids_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_movement_centroids"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Total_movement_centroids_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_movement_centroids"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Tail_base_movement_M1_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_tail_base"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Tail_base_movement_M1_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_tail_base"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Tail_base_movement_M1_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_tail_base"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Tail_base_movement_M2_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_tail_base"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Tail_base_movement_M2_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_tail_base"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Tail_base_movement_M2_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_tail_base"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Centroid_movement_M1_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_centroid"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Centroid_movement_M1_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_centroid"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Centroid_movement_M1_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_centroid"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Centroid_movement_M2_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_centroid"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Centroid_movement_M2_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_centroid"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Centroid_movement_M2_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_centroid"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Nose_movement_M1_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_nose"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Nose_movement_M1_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_nose"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Nose_movement_M1_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_1_nose"]
.rolling(int(window), min_periods=1)
.sum()
)
for window in self.roll_windows_values:
col_name = "Nose_movement_M2_median_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_nose"]
.rolling(int(window), min_periods=1)
.median()
)
col_name = "Nose_movement_M2_mean_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_nose"]
.rolling(int(window), min_periods=1)
.mean()
)
col_name = "Nose_movement_M2_sum_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Movement_mouse_2_nose"]
.rolling(int(window), min_periods=1)
.sum()
)
print("Calculating angles...")
self.out_data["Mouse_1_angle"] = self.angle3pt_vectorized(
data=self.out_data[
[
"Nose_1_x",
"Nose_1_y",
"Center_1_x",
"Center_1_y",
"Tail_base_1_x",
"Tail_base_1_y",
]
].values
)
self.out_data["Mouse_2_angle"] = self.angle3pt_vectorized(
data=self.out_data[
[
"Nose_2_x",
"Nose_2_y",
"Center_2_x",
"Center_2_y",
"Tail_base_2_x",
"Tail_base_2_y",
]
].values
)
self.out_data["Total_angle_both_mice"] = (
self.out_data["Mouse_1_angle"] + self.out_data["Mouse_2_angle"]
)
for window in self.roll_windows_values:
col_name = "Total_angle_both_mice_{}".format(str(window))
self.out_data[col_name] = (
self.out_data["Total_angle_both_mice"]
.rolling(int(window), min_periods=1)
.sum()
)
print("Calculating deviations...")
self.out_data["Total_movement_all_bodyparts_both_mice_deviation"] = (
self.out_data["Total_movement_all_bodyparts_both_mice"].mean()
- self.out_data["Total_movement_all_bodyparts_both_mice"]
)
self.out_data["Sum_euclid_distances_hull_deviation"] = (
self.out_data["Sum_euclidean_distance_hull_M1_M2"].mean()
- self.out_data["Sum_euclidean_distance_hull_M1_M2"]
)
self.out_data["M1_smallest_euclid_distances_hull_deviation"] = (
self.out_data["M1_hull_small_euclidean"].mean()
- self.out_data["M1_hull_small_euclidean"]
)
self.out_data["M1_largest_euclid_distances_hull_deviation"] = (
self.out_data["M1_hull_large_euclidean"].mean()
- self.out_data["M1_hull_large_euclidean"]
)
self.out_data["M1_mean_euclid_distances_hull_deviation"] = (
self.out_data["M1_hull_mean_euclidean"].mean()
- self.out_data["M1_hull_mean_euclidean"]
)
self.out_data["Centroid_distance_deviation"] = (
self.out_data["Centroid_distance"].mean()
- self.out_data["Centroid_distance"]
)
self.out_data["Total_angle_both_mice_deviation"] = (
self.out_data["Total_angle_both_mice"].mean()
- self.out_data["Total_angle_both_mice"]
)
self.out_data["Movement_mouse_1_deviation_centroid"] = (
self.out_data["Movement_mouse_1_centroid"].mean()
- self.out_data["Movement_mouse_1_centroid"]
)
self.out_data["Movement_mouse_2_deviation_centroid"] = (
self.out_data["Movement_mouse_2_centroid"].mean()
- self.out_data["Movement_mouse_2_centroid"]
)
self.out_data["Mouse_1_polygon_deviation"] = (
self.out_data["Mouse_1_poly_area"].mean()
- self.out_data["Mouse_1_poly_area"]
)
self.out_data["Mouse_2_polygon_deviation"] = (
self.out_data["Mouse_2_poly_area"].mean()
- self.out_data["Mouse_2_poly_area"]
)
for window in self.roll_windows_values:
col_name = "Total_movement_all_bodyparts_both_mice_mean_{}".format(
str(window)
)
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Sum_euclid_distances_hull_mean_{}".format(str(window))
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Mouse1_smallest_euclid_distances_mean_{}".format(
str(window)
)
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Mouse1_largest_euclid_distances_mean_{}".format(str(window))
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Mouse1_mean_euclid_distances_mean_{}".format(str(window))
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Movement_mean_{}".format(str(window))
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Distance_mean_{}".format(str(window))
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Total_angle_both_mice_{}".format(str(window))
deviation_col_name = col_name + "_deviation"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
print("Calculating percentile ranks...")
self.out_data["Movement_percentile_rank"] = self.out_data[
"Total_movement_centroids"
].rank(pct=True)
self.out_data["Distance_percentile_rank"] = self.out_data[
"Centroid_distance"
].rank(pct=True)
self.out_data["Movement_mouse_1_percentile_rank"] = self.out_data[
"Movement_mouse_1_centroid"
].rank(pct=True)
self.out_data["Movement_mouse_2_percentile_rank"] = self.out_data[
"Movement_mouse_1_centroid"
].rank(pct=True)
self.out_data["Movement_mouse_1_deviation_percentile_rank"] = self.out_data[
"Movement_mouse_1_deviation_centroid"
].rank(pct=True)
self.out_data["Movement_mouse_2_deviation_percentile_rank"] = self.out_data[
"Movement_mouse_2_deviation_centroid"
].rank(pct=True)
self.out_data["Centroid_distance_percentile_rank"] = self.out_data[
"Centroid_distance"
].rank(pct=True)
self.out_data["Centroid_distance_deviation_percentile_rank"] = (
self.out_data["Centroid_distance_deviation"].rank(pct=True)
)
for window in self.roll_windows_values:
col_name = "Total_movement_all_bodyparts_both_mice_mean_{}".format(
str(window)
)
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Sum_euclid_distances_hull_mean_{}".format(str(window))
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Mouse1_mean_euclid_distances_mean_{}".format(str(window))
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Mouse1_smallest_euclid_distances_mean_{}".format(
str(window)
)
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Mouse1_largest_euclid_distances_mean_{}".format(str(window))
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Movement_mean_{}".format(str(window))
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
for window in self.roll_windows_values:
col_name = "Distance_mean_{}".format(str(window))
deviation_col_name = col_name + "_percentile_rank"
self.out_data[deviation_col_name] = (
self.out_data[col_name].mean() - self.out_data[col_name]
)
print("Calculating path tortuosities...")
as_strided = np.lib.stride_tricks.as_strided
win_size = 3
centroid_lst_mouse1_x = as_strided(
self.out_data.Center_1_x,
(len(self.out_data) - (win_size - 1), win_size),
(self.out_data.Center_1_x.values.strides * 2),
)
centroid_lst_mouse1_y = as_strided(
self.out_data.Center_1_y,
(len(self.out_data) - (win_size - 1), win_size),
(self.out_data.Center_1_y.values.strides * 2),
)
centroid_lst_mouse2_x = as_strided(
self.out_data.Center_2_x,
(len(self.out_data) - (win_size - 1), win_size),
(self.out_data.Center_2_x.values.strides * 2),
)
centroid_lst_mouse2_y = as_strided(
self.out_data.Center_2_y,
(len(self.out_data) - (win_size - 1), win_size),
(self.out_data.Center_2_y.values.strides * 2),
)
for window in self.roll_windows_values:
start, end = 0, 0 + int(window)
tortuosities_results = defaultdict(list)
for frame in range(len(self.out_data)):
tortuosities_dict = defaultdict(list)
c_centroid_lst_mouse1_x, c_centroid_lst_mouse1_y = (
centroid_lst_mouse1_x[start:end],
centroid_lst_mouse1_y[start:end],
)
c_centroid_lst_mouse2_x, c_centroid_lst_mouse2_y = (
centroid_lst_mouse2_x[start:end],
centroid_lst_mouse2_y[start:end],
)
for frame_in_window in range(len(c_centroid_lst_mouse1_x)):
move_angle_mouse_1 = self.angle3pt(
c_centroid_lst_mouse1_x[frame_in_window][0],
c_centroid_lst_mouse1_y[frame_in_window][0],
c_centroid_lst_mouse1_x[frame_in_window][1],
c_centroid_lst_mouse1_y[frame_in_window][1],
c_centroid_lst_mouse1_x[frame_in_window][2],
c_centroid_lst_mouse1_y[frame_in_window][2],
)
move_angle_mouse_2 = self.angle3pt(
c_centroid_lst_mouse2_x[frame_in_window][0],
c_centroid_lst_mouse2_y[frame_in_window][0],
c_centroid_lst_mouse2_x[frame_in_window][1],
c_centroid_lst_mouse2_y[frame_in_window][1],
c_centroid_lst_mouse2_x[frame_in_window][2],
c_centroid_lst_mouse2_y[frame_in_window][2],
)
tortuosities_dict["Animal_1"].append(move_angle_mouse_1)
tortuosities_dict["Animal_2"].append(move_angle_mouse_2)
tortuosities_results["Animal_1"].append(
sum(tortuosities_dict["Animal_1"]) / (2 * math.pi)
)
tortuosities_results["Animal_2"].append(
sum(tortuosities_dict["Animal_2"]) / (2 * math.pi)
)
start += 1
end += 1
col_name = "Tortuosity_Mouse1_{}".format(str(window))
self.out_data[col_name] = tortuosities_results["Animal_1"]
print("Calculating pose probability scores...")
all_p_columns = self.mouse_2_p_headers + self.mouse_1_p_headers
self.out_data["Sum_probabilities"] = self.out_data[all_p_columns].sum(
axis=1
)
self.out_data["Sum_probabilities_deviation"] = (
self.out_data["Sum_probabilities"].mean()
- self.out_data["Sum_probabilities"]
)
self.out_data["Sum_probabilities_deviation_percentile_rank"] = (
self.out_data["Sum_probabilities_deviation"].rank(pct=True)
)
self.out_data["Sum_probabilities_percentile_rank"] = self.out_data[
"Sum_probabilities_deviation_percentile_rank"
].rank(pct=True)
results = pd.DataFrame(
self.count_values_in_range(
data=self.out_data.filter(all_p_columns).values,
ranges=np.array([[0.0, 0.1], [0.0, 0.5], [0.0, 0.75]]),
),
columns=[
"Low_prob_detections_0.1",
"Low_prob_detections_0.5",
"Low_prob_detections_0.75",
],
)
self.out_data = pd.concat([self.out_data, results], axis=1)
self.out_data = self.out_data.reset_index(drop=True).fillna(0)
save_path = os.path.join(
self.save_dir, self.video_name + "." + self.file_type
)
write_df(df=self.out_data, file_type=self.file_type, save_path=save_path)
video_timer.stop_timer()
print(
f"Feature extraction complete for {self.video_name} ({(file_cnt + 1)}/{len(self.files_found)} (elapsed time: {video_timer.elapsed_time_str}s)..."
)
self.timer.stop_timer()
stdout_success(
msg="All features extracted. Results stored in project_folder/csv/features_extracted directory",
elapsed_time=self.timer.elapsed_time_str,
)
# test = ExtractFeaturesFrom14bps(config_path='/Users/simon/Desktop/envs/troubleshooting/two_black_animals_14bp/project_folder/project_config.ini')
# test.run()