Module Contents


extract_multiple(reconstructions: List[Dict[str, Any]], feature_set: str, heavy_output_path: str, required_marks: Optional[List[str]] = None, only_marks: Optional[List[str]] = None, num_processes: Optional[int] = None, global_parameters: Optional[Dict[str, Any]] = None, output_table_path: Optional[str] = None) For each path in swc_paths, load the file into a morphology and (attempt
neuron_morphology.feature_extractor.__main__.extract_multiple(reconstructions: List[Dict[str, Any]], feature_set: str, heavy_output_path: str, required_marks: Optional[List[str]] = None, only_marks: Optional[List[str]] = None, num_processes: Optional[int] = None, global_parameters: Optional[Dict[str, Any]] = None, output_table_path: Optional[str] = None)

For each path in swc_paths, load the file into a morphology and (attempt to) extract each feature in the set specified by feature_set.

Because of how Windows handles multiprocessing, run_feature_extraction must be in another py file.

reconstructions : specify the reconstructions on which to compute features
feature_set : names the set of features for which calculation will be


heavy_output_path : write “heavy” outputs, such as arrays, to this h5 file
only_marks : names marks to which calculation will be restricted
required_marks : raise an exception if these named marks fail validation
num_processes : use this many cores in the multiprocessing pool.
global_parameters : a dictionary specifying cross-reconstruction


output_table_path : if not none, write a flattened table of features here
a dictionary whose keys are reconstruction identifers and whose values are

the outputs of run_feature_extraction for those reconstructions.