PANINIpy: Package of Algorithms for Nonparametric Inference with Networks in Python is a package designed for nonparametric inference with complex network data, with methods for identifying hubs in networks, regionalizing mobility or distributional data over spatial networks, clustering network populations, and constructing hypergraphs from temporal data among other features.
pip install paninipy
Identify MDL-optimal temporally contiguous partitions of event data between distinct node sets (e.g. users and products).
Utilizes method derived in “Inference of dynamic hypergraph representations in temporal interaction data” (Kirkley, 2024, https://arxiv.org/abs/2308.16546).
Generate synthetic network population datasets and perform clustering of observed network populations, multilayer network layers, or temporal networks.
Utilizes method derived in “Compressing network populations with modal networks reveals structural diversity” (Kirkley et al., 2023, https://arxiv.org/pdf/2209.13827).
Perform MDL-based regionalization on distributional (e.g. census) data over space.
Utilizes method derived in “Spatial regionalization as optimal data compression” (Kirkley, 2022, https://arxiv.org/pdf/2111.01813).
Identify hub nodes in a network using different information theoretic criteria.
Utilizes methods derived in “Identifying hubs in directed networks” (Kirkley, 2024, https://arxiv.org/pdf/2312.03347).
Perform community detection-based regionalization on network data.
Utilizes method derived in “Bayesian regionalization of urban mobility networks” (Morel-Balbi and Kirkley, 2024, https://journals.aps.org/prresearch/abstract/10.1103/PhysRevResearch.6.033307).
Infer global and local backbones of a network using the minimum description length principle .
Utilizes method derived in “Fast nonparametric inference of network backbones for graph sparsification” (Kirkley, 2024, https://arxiv.org/abs/2409.06417).
Detailed documentation for each module and function is available at the link below:
If you use PANINIpy in your research or projects, please cite it as follows:
For BibTeX users, include the following entry in the bibliography file:
@article{Kirkley2024,
doi = {10.21105/joss.07312},
url = {https://doi.org/10.21105/joss.07312},
year = {2024},
publisher = {The Open Journal},
volume = {9},
number = {103},
pages = {7312},
author = {Alec Kirkley and Baiyue He},
title = {PANINIpy: Package of Algorithms for Nonparametric Inference with Networks In Python},
journal = {Journal of Open Source Software}
}For APA users, cite as:
Kirkley, A., & He, B. (2024). PANINIpy: Package of Algorithms for Nonparametric Inference with Networks In Python. Journal of Open Source Software, 9(103), 7312, https://doi.org/10.21105/joss.07312
To reference PANINIpy, one can use either the BibTeX or APA format provided above. Please ensure that the citation includes the DOI and the link to the official article for clarity and accessibility.
For more details, visit the official article published in the Journal of Open Source Software.
Distributed under the MIT License. See LICENSE for more information.
The logo for this package was enhanced using Stable Diffusion model, an AI-based generative model created by Robin Rombach, Patrick Esser and contributors.
The model is released under the CreativeML Open RAIL-M License. For more details on the model and its licensing, refer to the following: