Reproducing the Analysis¶

Follow these steps to reproduce all results from the archived dataset.

1. Clone & install¶

git clone https://github.com/firmanhadi21/traffic-analyses.git
cd traffic-analyses

python3 -m venv venv
source venv/bin/activate        # Windows: venv\Scripts\activate
pip install -e ".[all]"

2. Download data from Zenodo¶

Download the dataset from doi:10.5281/zenodo.19211072 and unzip. Place the three traffic_*_output/ directories in the repository root:

traffic-analyses/
├── traffic_smg_output/   ← from Zenodo
├── traffic_bdg_output/   ← from Zenodo
└── traffic_jkt_output/   ← from Zenodo

3. Verify¶

traffic-pipeline --version
pytest tests/ -v              # requires pip install -e ".[dev]"

4. Run the pipeline¶

traffic-pipeline geostatistics    # Spatial statistics & hot-spot maps
traffic-pipeline bottleneck       # Road-capacity bottleneck analysis
traffic-pipeline poi              # POI-congestion density analysis
traffic-pipeline synthesis        # Temporal vs spatial comparison
traffic-pipeline multilevel       # Multilevel variance decomposition
traffic-pipeline markov           # LISA Markov transition analysis
traffic-pipeline speed-validation # Speed-based metric validation
traffic-pipeline h3-robustness    # H3 hexagonal MAUP robustness
traffic-pipeline positive-control # Positive control: FFS vs speed R²

Results are written to figures/ (PNG) and analysis_results/ (CSV).

Expected runtime¶

Stage	Approximate time
`geostatistics`	2–5 minutes
`bottleneck`	10–30 minutes (downloads OSMnx networks)
`poi`	5–15 minutes (downloads POI data)
`synthesis`	< 1 minute
`multilevel`	2–10 minutes
`markov`	5–15 minutes (LISA permutations)
`speed-validation`	1–3 minutes
`h3-robustness`	2–5 minutes