repic.utils.score_detections
Calaculates performance metrics of particle picking algorithm by comparing output to the normative (final particle sets from EMPIAR)
Attributes
argparse parse_args() object |
Functions
|
Creates segmanetation maps of particle picking algorithm and normative particle bounding boxes, then calculates performance metrics (precision, recall, F1-score, and positive fraction of particles) |
Module Contents
- repic.utils.score_detections.get_segmentation_scores(norm_boxes, pckr_boxes, conf_thresh=None, mrc_w=None, mrc_h=None)
Creates segmanetation maps of particle picking algorithm and normative particle bounding boxes, then calculates performance metrics (precision, recall, F1-score, and positive fraction of particles)
- Parameters:
norm_boxes (list) – list of normative particle bounding boxes
pckr_boxes (list) – list of particle picking algorithm bounding boxes
- Keyword Arguments:
conf_thresh (float or None) – false positive filtering threshold of particle picking algorithm
mrc_w (int or None) – micrograph width (pixels)
mrc_h (int or None) – micrograph height (pixels)
- Returns:
precision, recall, F1-score, and positive fraction of pixels
- Return type:
float, float, float, float
- repic.utils.score_detections.parser
argparse parse_args() object
- Type:
obj