Inclusive Mobility in Chiang Mai

Description

Version 2 expands and revises the dataset accompanying the study of walking and assisted-wheelchair mobility in Chiang Mai, Thailand. It contains a de-identified merged analysis dataset and analytical code supporting the study “Context-dependent associations between pedestrian exposure and traversal speed: A state-space analysis of walking and assisted-wheelchair mobility.” The dataset covers 18 traversal tracks across three urban corridors (Nimmanhemin, Ang Kaew, and the Temples corridor), two movement modes, and three occasions. The repository includes R and Stan code for reproducing the Bayesian state-space analysis, together with R and Python scripts documenting the upstream GPS–video integration and YOLOv8 object-detection procedures. Raw street-level video and intermediate object-detection files are not publicly released. See README.md for the scope of reproducibility, variable definitions, and execution order.

Steps to reproduce

Download and extract the repository files. The statistical analysis can be reproduced beginning with data/temples_nimman_angkaew_merged.csv. Run 01_prepare_model_data.R, 02_fit_model.R, 03_check_results.R, and 04_draw_figures.R in numerical order. The Stan model used by the fitting script is stan/ssm_place_interaction_roll5.stan. The scripts 00_build_merged_data.R and yolov8_object_detection.py document upstream processing procedures; their original input files are not publicly released. See README.md for details.

Institutions

• Nagasaki University

  Nagasaki, Nagasaki
• Chiang Mai University

  Chiang Mai, Chiang Mai

Categories

Funders

• Chiang Mai University

  Chiang Mai

  Grant ID: 67215E146C-0023M

Licence