Cross-Condition Network Analysis

Overview

This repository contains the analysis code for the paper:

Blue-Enriched White Light and Within-Subsystem Integration in Preschool Children: An Exploratory Network Analysis Dissociating Univariate and Topological Sensitivity

Yankang Jiang ^a, Dingheng Mai ^b, Xiaotong Chen ^c, Shuxin Wang ^d, Junxian Liang ^e, Yupeng Shen ^{a, *}

_{^a Sports Engineering Center, School of Physical Education and Sports Science, South China Normal University, Guangzhou, China
^b Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States
^c School of Physical Education, Guangzhou Business School, Guangzhou, China
^d College of Physical Education and Health, Guangdong Polytechnic Normal University, Guangzhou, China
^e Center for Cognitive and Brain Sciences, Institute of Collaborative Innovation, University of Macau, Taipa, Macau, China}

We implement a correlation-based network analysis comparing two experimental conditions (CWL vs. BWL) using sign-flip permutation inference on global, node-level, and edge-level network metrics. The entire analysis is fully reproducible from raw data to final figures in a single Jupyter notebook.

Method Summary

Step	Description
Network construction	Spearman correlation matrix; edges thresholded at p < .05
Node metric	Weighted strength
Global metrics	Density, mean edge weight, global efficiency, weighted average clustering coefficient, weighted average shortest path length, modularity
Inference	10,000 sign-flip permutations (paired design)
Multiple-comparison correction	Benjamini–Hochberg FDR

Note: All analysis parameters (e.g., significance threshold, number of permutations, FDR level) are defined in the first code cell of the Jupyter notebook and can be modified there.

Key Results

Figure 1 — Condition-Specific Networks & Strength Change

(A) CWL network. (B) BWL network. (C) Difference network (BWL − CWL). (D) Node strength change.

Figure 2 — Forest Plot & Strength Rank Changes

(A) Node strength ± 95% permutation CI. (B) Bump chart of strength rank migration.

Repository Structure

.
├── data/ # Raw data (not included in repo)
├── output/figures/ # Generated figures
├── CAN_network_analysis.ipynb # Main analysis notebook
├── requirements.txt # Python dependencies
├── LICENSE.txt # MIT License
└── README.md # This file

Installation

git clone https://github.com/YKang97/CAN_network_analysis.git
cd CAN_network_analysis
pip install -r requirements.txt

How to Reproduce

1. Clone this repository
2. Install dependencies: pip install -r requirements.txt
3. Place your data file at ./data/CAN.xlsx
4. Open CAN_network_analysis.ipynb in Jupyter
5. Run all cells sequentially (Cell 1 → Cell 6)

Data Availability

Raw data are available upon reasonable request from the corresponding author. The data file is not included in this repository to protect participant privacy.

Citation

If you use this code, please cite:

@article{Jiang2026,
  title   = {Blue-Enriched White Light and Within-Subsystem Integration
             in Preschool Children: An Exploratory Network Analysis
             Dissociating Univariate and Topological Sensitivity},
  author  = {Jiang, Yankang and Mai, Dingheng and Chen, Xiaotong
             and Liang, Junxian and Shen, Yupeng},
  journal = {},
  year    = {2026},
  doi     = {10.5281/zenodo.19459881}
}

License

This project is licensed under the MIT License. See LICENSE.