Research validation (Horizon 9)

ts2net ships literature fixtures, bundled UCR datasets, and reproducibility manifests.

Datasets

• ts2net.datasets — synthetic_causal, synthetic_classification, synthetic_anomaly, synthetic_regime, spain_meters_summary

• Bundled UCR archives under ts2net/datasets/data/ucr/

• ts2net.datasets.load_ucr(), ts2net.datasets.run_ucr_benchmark()

Fixtures

Run all checks:

python benchmarks/run_validation.py
TS2NET_CI_SMOKE=1 python benchmarks/run_validation.py

Recorded UCR baselines live in ts2net/datasets/data/ucr_baselines.json.

Outputs

• benchmarks/results/validation_manifest.json

• benchmarks/results/ucr_benchmark.json