Benchmarks

Validation Dashboard#

Track accuracy and seed stability across releases. Per-metric charts show how closely simulation results match calibration targets and how consistently random seeds produce valid outcomes.

Benchmark List#

Flat table of all validation metrics across scenarios. Sortable by score, stability, or weight, filterable by scenario and metric group.

Regressions#

All detected accuracy regressions across releases. Filterable by scenario, metric group, and severity threshold.

Performance (ASV)#

Performance benchmarks tracked with ASV (Airspeed Velocity).

View the latest benchmark results.

What is Benchmarked#

• Baseline simulation throughput (periods/second)
• Economy scaling (100–500 firms)
• Extension overhead (Growth+, Buffer-Stock, Taxation)

Running Locally#

git clone https://github.com/kganitis/bam-engine.git
cd bam-engine/asv_benchmarks
asv run
asv publish
asv preview