simba-uw-tf-dev 4.7.1__tar.gz → 4.7.2__tar.gz

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: simba_uw_tf_dev
-Version: 4.7.1
+Version: 4.7.2
 Summary: Toolkit for computer classification and analysis of behaviors in experimental animals
 Home-page: https://github.com/sgoldenlab/simba
 Author: Simon Nilsson, Jia Jie Choong, Sophia Hwang

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/setup.py

@@ -29,7 +29,7 @@ with open(GPU_REQUIREMENTS_PATH, "r") as f:
 # Setup configuration
 setuptools.setup(
     name="simba_uw_tf_dev",
-    version="4.7.1",
+    version="4.7.2",
     author="Simon Nilsson, Jia Jie Choong, Sophia Hwang",
     author_email="sronilsson@gmail.com",
     description="Toolkit for computer classification and analysis of behaviors in experimental animals",

simba_uw_tf_dev-4.7.2/simba/assets/lookups/yolo_schematics/yolo_mitra.csv

@@ -0,0 +1,9 @@
+left_ear
+right_ear
+nose
+left_side
+right_side
+tail_base
+center
+tail_center
+tail_tip

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/mixins/train_model_mixin.py

@@ -1070,10 +1070,7 @@ class TrainModelMixin(object):
                 MissingUserInputWarning(msg=f'Skipping {str(config.get("SML settings", "target_name_" + str(n + 1)))} classifier analysis: missing information (e.g., no discrimination threshold and/or minimum bout set in the project_config.ini',source=self.__class__.__name__)
 
         if len(model_dict.keys()) == 0:
-            raise NoDataError(
-                msg=f"There are no models with accurate data specified in the RUN MODELS menu. Specify the model information to SimBA RUN MODELS menu to use them to analyze videos",
-                source=self.get_model_info.__name__,
-            )
+            raise NoDataError(msg=f"There are no models with accurate data specified in the RUN MODELS menu. Specify the model information to SimBA RUN MODELS menu to use them to analyze videos. PLease check the model paths, thresholds, and minimum bout lengths.", source=self.get_model_info.__name__)
         else:
             return model_dict
 

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/model/inference_batch.py

@@ -101,7 +101,7 @@ class InferenceBatch(TrainModelMixin, ConfigReader):
             video_timer.stop_timer()
             print(f"Predictions created for {file_name} (frame count: {len(in_df)}, elapsed time: {video_timer.elapsed_time_str}) ...")
         self.timer.stop_timer()
-        stdout_success(msg=f"Machine predictions complete. Files saved in {self.save_dir} directory", elapsed_time=self.timer.elapsed_time_str, source=self.__class__.__name__)
+        stdout_success(msg=f"Machine predictions complete for {len(self.feature_file_paths)} file(s). Files saved in {self.save_dir} directory", elapsed_time=self.timer.elapsed_time_str, source=self.__class__.__name__)
 
 if __name__ == "__main__" and not hasattr(sys, 'ps1'):
     parser = argparse.ArgumentParser(description="Perform classifications according to rules defined in SImAB project_config.ini.")

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/third_party_label_appenders/transform/simba_to_yolo.py

@@ -117,12 +117,15 @@ class SimBA2Yolo:
         annotations, timer, body_part_headers = [], SimbaTimer(start=True), []
         for file_cnt, video_name in enumerate(self.data_w_video):
             data = read_df(file_path=self.data_paths[video_name], file_type=self.config.file_type)
+            data.columns = [x.lower() for x in list(data.columns)]
+            bp_header_names = [x.lower() for x in self.config.bp_headers]
             check_valid_dataframe(df=data, source=f'{self.__class__.__name__} {self.data_paths[video_name]}', valid_dtypes=Formats.NUMERIC_DTYPES.value)
             video_path = self.video_paths[video_name]
             check_video_and_data_frm_count_align(video=video_path, data=data, name=self.data_paths[video_name], raise_error=True)
             p_data = data[data.columns[list(data.columns.str.endswith('_p'))]]
             data = data.loc[:, ~data.columns.str.endswith('_p')].reset_index(drop=True)
             data = data.iloc[(p_data[(p_data > self.threshold).all(axis=1)].index)]
+            data = data[[x for x in bp_header_names if not x.endswith('_p')]]
             body_part_headers = data.columns
             data['video'], frm_cnt = video_name, len(data)
             if self.sample_size is None:
@@ -155,7 +158,7 @@ class SimBA2Yolo:
             if frm_idx in train_idx:
                 img_save_path, lbl_save_path = os.path.join(self.img_train_dir, f'{file_name}.png'), os.path.join(self.lbl_train_dir, f'{file_name}.txt')
             else:
-                img_save_path, lbl_save_path = os.path.join(self.img_train_dir, f'{file_name}.png'), os.path.join(self.lb_val_dir, f'{file_name}.txt')
+                img_save_path, lbl_save_path = os.path.join(self.img_val_dir, f'{file_name}.png'), os.path.join(self.lb_val_dir, f'{file_name}.txt')
             img = read_frm_of_video(video_path=vid_path, frame_index=frm_idx, greyscale=self.greyscale, clahe=self.clahe)
             img_h, img_w = img.shape[0], img.shape[1]
             keypoints_with_id = {}
@@ -181,8 +184,8 @@ class SimBA2Yolo:
         timer.stop_timer()
         stdout_success(msg=f'YOLO formated data saved in {self.save_dir} directory', source=self.__class__.__name__, elapsed_time=timer.elapsed_time_str)
 
-
-# SAVE_DIR = r'D:\troubleshooting\mitra\mitra_yolo'
-# CONFIG_PATH = r"C:\troubleshooting\mitra\project_folder\project_config.ini"
-# runner = SimBA2Yolo(config_path=CONFIG_PATH, save_dir=SAVE_DIR, sample_size=10, verbose=True, names=('animal_1',))
+#
+# SAVE_DIR = r'E:\troubleshooting\mitra\yolo_0126\yolo_train_0126'
+# CONFIG_PATH = r"E:\troubleshooting\mitra\project_folder\project_config.ini"
+# runner = SimBA2Yolo(config_path=CONFIG_PATH, save_dir=SAVE_DIR, sample_size=50, verbose=True, names=('animal_1',), threshold=0.5)
 # runner.run()

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/ui/pop_ups/run_machine_models_popup.py

@@ -8,7 +8,7 @@ from simba.mixins.config_reader import ConfigReader
 from simba.mixins.pop_up_mixin import PopUpMixin
 from simba.model.inference_batch import InferenceBatch
 from simba.ui.tkinter_functions import (CreateLabelFrameWithIcon, Entry_Box,
-                                        FileSelect, SimbaButton, SimBADropDown,
+                                        FileSelect, SimbaButton, SimBASeperator,
                                         SimBALabel)
 from simba.utils.checks import (check_file_exist_and_readable, check_float,
                                 check_int)
@@ -31,34 +31,34 @@ class RunMachineModelsPopUp(PopUpMixin, ConfigReader):
     def __init__(self, config_path: Union[str, os.PathLike]):
         ConfigReader.__init__(self, config_path=config_path, read_video_info=False)
         PopUpMixin.__init__(self, title="SET MODEL PARAMETERS", icon='equation_small')
+        padx, self.config_path = (0, 25), config_path
         self.clf_table_frm = CreateLabelFrameWithIcon(parent=self.main_frm, header="SETTINGS", icon_name=Keys.DOCUMENTATION.value, icon_link=Links.SET_RUN_ML_PARAMETERS.value)
-        clf_header = SimBALabel(parent=self.clf_table_frm, txt="CLASSIFIER", font=Formats.FONT_HEADER.value)
-        mdl_path_header = SimBALabel(parent=self.clf_table_frm, txt="MODEL PATH (.SAV)", font=Formats.FONT_HEADER.value)
-        threshold_header = SimBALabel(parent=self.clf_table_frm, txt="THRESHOLD (0.0 - 1.0)", font=Formats.FONT_HEADER.value)
-        min_bout_header = SimBALabel(parent=self.clf_table_frm, txt="MINIMUM BOUT LENGTH (MS)", font=Formats.FONT_HEADER.value)
-        clf_header.grid(row=0, column=0, sticky=W, padx=(0, 10))
-        mdl_path_header.grid(row=0, column=1, sticky=NW)
-        threshold_header.grid(row=0, column=2, sticky=NW)
-        min_bout_header.grid(row=0, column=3, sticky=NW)
+        clf_header = SimBALabel(parent=self.clf_table_frm, txt="CLASSIFIER", font=Formats.FONT_HEADER.value, img='label')
+        mdl_path_header = SimBALabel(parent=self.clf_table_frm, txt="MODEL PATH (.SAV)", font=Formats.FONT_HEADER.value, img='file_type', justify='center')
+        threshold_header = SimBALabel(parent=self.clf_table_frm, txt="THRESHOLD (0.0 - 1.0)", font=Formats.FONT_HEADER.value, img='threshold', justify='center')
+        min_bout_header = SimBALabel(parent=self.clf_table_frm, txt="MINIMUM BOUT LENGTH (MS)", font=Formats.FONT_HEADER.value, img='timer_2', justify='center')
+        clf_header.grid(row=0, column=0, sticky=NW, padx=padx)
+        mdl_path_header.grid(row=0, column=1, sticky=NW, padx=padx)
+        threshold_header.grid(row=0, column=2, sticky=NW, padx=padx)
+        min_bout_header.grid(row=0, column=3, sticky=NW, padx=padx)
+
+        seperator = SimBASeperator(parent=self.clf_table_frm, color='grey', orient='horizontal', borderwidth=1)
+        seperator.grid(row=1, column=0, columnspan=4, rowspan=1, sticky="ew", pady=(0, 10))
 
         self.clf_data = {}
         for clf_cnt, clf_name in enumerate(self.clf_names):
             self.clf_data[clf_name] = {}
-            SimBALabel(parent=self.clf_table_frm, txt=clf_name, font=Formats.FONT_REGULAR_ITALICS.value).grid(row=clf_cnt + 1, column=0, sticky=W, padx=(0, 10))
+            SimBALabel(parent=self.clf_table_frm, txt=clf_name, font=Formats.FONT_REGULAR_ITALICS.value).grid(row=clf_cnt + 2, column=0, sticky=W, padx=padx)
             mdl_path = read_config_entry(config=self.config, section=ConfigKey.SML_SETTINGS.value, option=f"model_path_{clf_cnt + 1}", default_value='Select model (.sav) file', data_type=Dtypes.STR.value)
             self.clf_data[clf_name][PATH] = FileSelect(self.clf_table_frm, title="Select model (.sav) file", initialdir=self.project_path, file_types=[("SimBA Classifier", "*.sav")], initial_path=mdl_path)
             threshold = read_config_entry(config=self.config, section=ConfigKey.THRESHOLD_SETTINGS.value, option=f"threshold_{clf_cnt + 1}", default_value='', data_type=Dtypes.STR.value)
-            self.clf_data[clf_name][THRESHOLD] = Entry_Box(parent=self.clf_table_frm, fileDescription='', labelwidth=0, entry_box_width=20, value=threshold)
+            self.clf_data[clf_name][THRESHOLD] = Entry_Box(parent=self.clf_table_frm, fileDescription='', labelwidth=0, entry_box_width=20, value=threshold, justify='center')
             bout_length = read_config_entry(config=self.config, section=ConfigKey.MIN_BOUT_LENGTH.value, option=f"min_bout_{clf_cnt + 1}", default_value='', data_type=Dtypes.STR.value)
-            self.clf_data[clf_name][MIN_BOUT] = Entry_Box(parent=self.clf_table_frm, fileDescription='', labelwidth=0, entry_box_width=20, value=bout_length)
-            self.clf_data[clf_name][PATH].grid(row=clf_cnt + 1, column=1, sticky=NW, padx=(0, 5))
-            self.clf_data[clf_name][THRESHOLD].grid(row=clf_cnt + 1, column=2, sticky=NW, padx=(0, 5))
-            self.clf_data[clf_name][MIN_BOUT].grid(row=clf_cnt + 1, column=3, sticky=NW, padx=(0, 5))
+            self.clf_data[clf_name][MIN_BOUT] = Entry_Box(parent=self.clf_table_frm, fileDescription='', labelwidth=0, entry_box_width=20, value=bout_length, justify='center')
+            self.clf_data[clf_name][PATH].grid(row=clf_cnt + 2, column=1, sticky=NW, padx=padx)
+            self.clf_data[clf_name][THRESHOLD].grid(row=clf_cnt + 2, column=2, sticky=NW, padx=padx)
+            self.clf_data[clf_name][MIN_BOUT].grid(row=clf_cnt + 2, column=3, sticky=NW, padx=padx)
         self.clf_table_frm.grid(row=0, sticky=W, pady=5, padx=5)
-        # self.runtime_settings_frm = CreateLabelFrameWithIcon(parent=self.main_frm, header="RUNTIME SETTINGS", icon_name='run', icon_link=Links.SET_RUN_ML_PARAMETERS.value)
-        # self.core_cnt_dropdown = SimBADropDown(parent=self.runtime_settings_frm, dropdown_options=CORE_CNT_OPTIONS, label='CPU CORE COUNT:', label_width=30, dropdown_width=20, value=int(find_core_cnt()[0]/3))
-        # self.runtime_settings_frm.grid(row=1, sticky=W, pady=5, padx=5)
-        # self.core_cnt_dropdown.grid(row=0, sticky=W, pady=5, padx=5)
         run_frm = CreateLabelFrameWithIcon(parent=self.main_frm, header=f"ANALYZE {len(self.feature_file_paths)} FILES(S)", icon_name='rocket')
         run_btn = SimbaButton(parent=run_frm, txt="RUN", img='rocket', txt_clr='red', font=Formats.FONT_REGULAR.value, hover_font=Formats.FONT_REGULAR.value, cmd=self.run)
         run_frm.grid(row=2, sticky=W, pady=5, padx=5)
@@ -95,7 +95,7 @@ class RunMachineModelsPopUp(PopUpMixin, ConfigReader):
         with open(self.config_path, "w") as f:
             self.config.write(f)
 
-        stdout_success(msg="Model paths/settings saved in project_config.ini", source=self.__class__.__name__)
+        stdout_success(msg=f"Model paths/settings saved in project_config.ini ({self.config_path})", source=self.__class__.__name__)
 
         if len(self.feature_file_paths) == 0:
             raise NoDataError(msg=f'Cannot run machine model predictions: No data files found in {self.features_dir} directory', source=self.__class__.__name__)
@@ -103,4 +103,4 @@ class RunMachineModelsPopUp(PopUpMixin, ConfigReader):
         inferencer = InferenceBatch(config_path=self.config_path, features_dir=None, save_dir=None, minimum_bout_length=None)
         inferencer.run()
 
-#_ = RunMachineModelsPopUp(config_path=r"D:\troubleshooting\mitra\project_folder\project_config.ini")
+#_ = RunMachineModelsPopUp(config_path=r"E:\troubleshooting\mitra_emergence\project_folder\project_config.ini")

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/ui/pop_ups/simba_to_yolo_keypoints_popup.py

@@ -63,7 +63,7 @@ class SimBA2YoloKeypointsPopUp(PopUpMixin):
         animal_names = list(config.animal_bp_dict.keys())
         bps = [x[:-2] for x in config.animal_bp_dict[animal_names[0]]['X_bps']]
         flip_idx = get_yolo_keypoint_flip_idx(x=bps)
-        map_dict = {c: k for c, k in enumerate(animal_names)}
+        #map_dict = {c: k for c, k in enumerate(animal_names)}
         bp_id_idx = None
         if len(animal_names) > 1:
             bp_id_idx = get_yolo_keypoint_bp_id_idx(animal_bp_dict=config.animal_bp_dict)
@@ -85,7 +85,7 @@ class SimBA2YoloKeypointsPopUp(PopUpMixin):
                             greyscale=grey,
                             padding=padding,
                             flip_idx=flip_idx,
-                            names=map_dict,
+                            names=tuple(animal_names),
                             sample_size=sample_size,
                             bp_id_idx=bp_id_idx,
                             clahe=clahe)

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/ui/pop_ups/yolo_inference_popup.py

@@ -30,7 +30,7 @@ SMOOTHING_OPTIONS = ['None', 50, 100, 200, 300, 400, 500]
 
 YOLO_FORMATS = Options.VALID_YOLO_FORMATS.value + ['None']
 devices = ['CPU']
-THRESHOLD_OPTIONS = list(np.arange(0.1, 1.1, 0.1).astype(np.float32))
+THRESHOLD_OPTIONS = list(np.arange(0.05, 1.05, 0.05).astype(np.float32))
 
 simba_dir = os.path.dirname(simba.__file__)
 YOLO_SCHEMATICS_DIR = os.path.join(simba_dir, Paths.YOLO_SCHEMATICS_DIR.value)

{simba_uw_tf_dev-4.7.1 → simba_uw_tf_dev-4.7.2}/simba/ui/pop_ups/yolo_pose_train_popup.py

@@ -33,7 +33,7 @@ class YOLOPoseTrainPopUP(PopUpMixin):
         PopUpMixin.__init__(self, title="TRAIN YOLO POSE ESTIMATION MODEL", icon='ultralytics_2')
         settings_frm = CreateLabelFrameWithIcon(parent=self.main_frm, header="SETTINGS", icon_name='settings')
         devices.extend([f'{x} : {y["model"]}' for x, y in gpus.items()])
-        self.yolo_map_path = FileSelect(parent=settings_frm, fileDescription='YOLO MAP FILE (YAML):', lblwidth=35, entry_width=45, file_types=[("YOLO MODEL FILE", Options.ALL_YOLO_MODEL_FORMAT_STR_OPTIONS.value)], lbl_icon='file')
+        self.yolo_map_path = FileSelect(parent=settings_frm, fileDescription='YOLO MAP FILE (YAML):', lblwidth=35, entry_width=45, file_types=[("YOLO MODEL FILE", ".yaml")], lbl_icon='file')
         self.save_dir = FolderSelect(settings_frm, folderDescription="SAVE DIRECTORY:", lblwidth=35, entry_width=45, lbl_icon='save')
         self.weights_path = FileSelect(parent=settings_frm, fileDescription='INITIAL WEIGHT FILE (E.G., .PT):', lblwidth=35, entry_width=45, lbl_icon='file')