Skip to content

SpectraNorm

A Python package for spectral normative modeling of high-dimensional data.

PyPI PyPI - Python Version PyPI - License Coookiecutter - Wolt

Installation

You can install the package using pip:

pip install spectranorm --upgrade

Installation typically takes less than 1 minute on a standard laptop with an internet connection.

Requirements

SpectraNorm is implemented in Python and depends on several standard scientific Python libraries, including:

  • numpy
  • scipy
  • pandas
  • nibabel
  • joblib
  • pymc

All dependencies are automatically installed via pip.

While SpectraNorm should work in any Python 3.10+ environment, it has been specifically tested on a conda environment on a Linux machine running Python 3.12.

Getting Started

Check out the tutorials to get started with using SpectraNorm.

API Reference

The API reference provides detailed documentation of several functions and classes available in the package.

Dig Deeper

For more in-depth information about the underlying theory and example uses of spectral normative modeling, check out this GitHub repository which hosts several notebooks that implement spectral normative modeling of cortical thickness phenotypes on a large scale population-wide imaging biobank. You may also be interested in the [original paper][link to be added] describing the method and its application to brain imaging data.