Datasets:
| license: cc-by-4.0 | |
| tags: [causal-inference, time-series, benchmark, dotime] | |
| # dot-Identifiability-v1 | |
| A frozen evaluation suite from **DoTime**. | |
| - Episodes: 10800 | |
| - Schema: parquet shards + `manifest.json` (md5-checksummed), Croissant metadata. | |
| - Load with: | |
| ```python | |
| from dotime.benchmarks import load_benchmark | |
| suite = load_benchmark("dot-Identifiability-v1") # pulls this repo at tag v1.0.0 | |
| ``` | |
| Generated reproducibly by `scripts/build_release.py`. Zenodo DOI is the citable | |
| archive of record. | |