moseq2_app.roi package
Submodules
ROI - Controller Module
Interactive ROI detection and extraction preview functionalities.
- class moseq2_app.roi.controller.InteractiveExtractionViewer(data_path, flipped=False)
Bases:
object- clear_on_click(b=None)
Clear the cell output
Args: b (button click)
- get_extraction(input_file)
Returns a div containing a video object to display.
Args: input_file (str): Path to session extraction video to view.
ROI - Validation Module
The module contains extraction validation functions that test extractions’ scalar values timestamps, and position heatmaps.
- moseq2_app.roi.validation.check_timestamp_error_percentage(timestamps, fps=30, scaling_factor=1000)
Return the proportion of dropped frames relative to the respective recorded timestamps and frames per second.
Args: timestamps (numpy.array): Session’s recorded timestamp array. fps (int): Frames per second scaling_factor (float): factor to divide timestamps by to convert timestamp milliseconds into seconds.
Returns: percentError (float): Percentage of frames that were dropped/missed during acquisition.
- moseq2_app.roi.validation.count_frames_with_small_areas(scalar_df)
Count the number of frames where the mouse area is smaller than 2 standard deviations of all mouse areas.
Args: scalar_df (pd.DataFrame): dataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df)
Returns: corrupt_frames (int): Number of frames where the recorded mouse area is too small
- moseq2_app.roi.validation.count_missing_mouse_frames(scalar_df)
Count the number of frames where the mouse is not found.
Args: scalar_df (pd.DataFrame): dataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df)
Returns: missing_mouse_frames (int): Number of frames with recorded mouse area ~= 0
- moseq2_app.roi.validation.count_nan_rows(scalar_df)
Count the number of rows with NaN scalar values.
Args: scalar_df (pd.DataFrame): dataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df)
Returns: n_missing_frames (int): Number of frames with NaN computed scalar values.
- moseq2_app.roi.validation.count_stationary_frames(scalar_df)
Count the number of frames where mouse is not moving.
Args: scalar_df (pd.DataFrame): dataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df)
Returns: motionless_frames (int): Number of frames where the mouse is not moving
- moseq2_app.roi.validation.get_scalar_anomaly_sessions(scalar_df, status_dicts)
Detect outlier sessions using an EllipticEnvelope model based on a subset of their mean scalar values.
Args: scalar_df (pd.DataFrame): dataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df) status_dicts (dict): stacked dictionary object containing all the sessions’ flag status dicts.
Returns: status_dicts (dict): stacked dictionary object containing updated scalar_anomaly flags.
- moseq2_app.roi.validation.get_scalar_df(path_dict)
Compute a scalar dataframe that contains all the extracted sessions recorded scalar values along with their metadata.
Args: path_dict (dict): dictionary of session folder names paired with their extraction paths
Returns: scalar_df (pd.DataFrame): Ddataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df)
- moseq2_app.roi.validation.make_session_status_dicts(paths)
Return the flag status dicts for all the found completed extracted sessions. Additionally performs dropped frames test on all sessions.
Args: paths (dict): path dict of session names paired wit their mp4 paths.
Returns: status_dicts (dict): stacked dictionary object containing all the sessions’ flag status dicts.
- moseq2_app.roi.validation.print_validation_results(scalar_df, status_dicts)
Display all the outlier sessions flag names and values. Additionally plots the flagged position heatmap.
Args: anomaly_dict (dict): Dict object containing specific session flags to print
- moseq2_app.roi.validation.run_validation_tests(scalar_df, status_dicts)
run all the available extraction validation tests and updates the status_dicts flags accordingly.
Args: scalar_df (pd.DataFrame): dataframe that contains frame by frame scalar and syllable data (moseq_df/scalar_df) status_dicts (dict): stacked dictionary object containing all the sessions’ flag status dicts.
Returns: status_dicts (dict): stacked dictionary object containing all the sessions’ updated flag status dicts.
ROI - View Module
Interactive ROI/Extraction Bokeh visualization functions.
- moseq2_app.roi.view.bokeh_plot_helper(bk_fig, image)
create the Bokeh image gylphs in the created canvases/figures.
Args: bk_fig (Bokeh figure): figure canvas to draw image/glyph on image (2D np.array): image to draw.
- moseq2_app.roi.view.show_extraction(input_file, video_file)
display manually triggered extraction.s
Args: input_file (str): session name to display. video_file (str): path to video to display
Returns: output (ipywidgets.Output widget): HTML canvas where the video is being displayed.