Data and analysis code to reproduce the main results for the manuscript

Data and analysis code to reproduce the main results for the manuscript: Baseline Physical Activity Phenotype Modulates Acute and Chronic Glycemic Responses to Mixed Exercise in Type 2 Diabetes: A Randomized Crossover Trial This repository contains (i) de-identified, analysis-ready datasets and (ii) Python scripts used to generate the key figures and associated statistics (Figures 2–4) reported in the manuscript. What is included De-identified, analysis-ready datasets in data/raw/ Figure reproduction scripts in scripts/ License file (MIT) Repository structure data/raw/ pre_cgm.xlsx — CGM-derived outcomes for the acute/baseline window (analysis-ready) post_cgm.xlsx — CGM-derived outcomes for the chronic/follow-up window (analysis-ready) daily_activity.xlsx — free-living physical activity summaries used to define baseline PA phenotype biochemical.xlsx — biochemical indicators (pre/post) used as outcomes and/or covariates scripts/ 02_Figure_2.py 03_Figure_3.py 04_Figure_4.py outputs/ (created after running scripts) outputs/figures/ outputs/tables/ Data notes (privacy and scope) The shared files are analysis-ready datasets. Raw device exports (high-frequency CGM time series and raw accelerometer epoch-level files) are not included to reduce re-identification risk. All participant identifiers are de-identified (e.g., study IDs only). Variable definitions, units, and derived-variable rules should be documented in the script headers. If you plan to expand this repo, it is recommended to add a dedicated codebook in data/codebook/. Requirements Python 3 is required. The scripts typically depend on: pandas, numpy scipy statsmodels matplotlib (and optionally seaborn) Recommended: add one of the following files to make installation “one-step”: requirements.txt (pip) or environment.yml (conda) Example (pip): pip install pandas numpy scipy statsmodels matplotlib seaborn Contact For questions/issues, please open a GitHub Issue or contact:Bin Gai — [e-mail]