Skip to content

PyNeval, a toolbox for quantative analysis on neuron reconstruction result.

License

Notifications You must be signed in to change notification settings

CSDLLab/PyNeval

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PyNeval

docs-badge

Introduction

PyNeval is a Python package for evaluating the qualities of neuron reconstructions in the SWC format. It outputs quality scores of one or more test models by comparing them to a gold standard model. The scores also depend on which metric is specified because PyNeval supports several metric options.

Here is the simplest command line interface to run PyNeval:

pyneval --test <test_swc_path> --gold <gold_swc_path> --metric <metric>

test_swc_path is the file path to a test model, which is often produced by an automatic reconstruction method, and gold_swc_path is the file path to the gold standard model, which is typically created by manual editing. metric is the name of a quality metric, which can be

  • length: Length metric for measuring the overlapping ratio between two models by matching line segments.
  • ssd: SSD metric for measuring the overlapping ratio between two models by matching resampled nodes.
  • diadem: DIADEM metric for measuring the amount of paths that contribute to topological similarity between two models.
  • cn: Critical node metric for measuring the topological similarity between two models by matching topologically critical nodes.

Installation

pip

pip install pyneval

conda

conda install pyneval -c csdllab -c conda-forge

source code

git clone https://github.com/bennieHan/pyNeval.git
cd PyNeval
python setup.py install

Example

Once PyNeval is installed successfully, you can test it with the demo data included in the source code.

pyneval --gold data/demo/gold.swc --test data/demo/test.swc --metric ssd