The official ORCESTRA dataset is distributed via IPFS. The latest version is always available via the IPNS name:
latest.orcestra-campaign.org
Technically, IPFS is not restricted to any particular data format. However, certain data formats allow the user to access data directly via http(s). One of these data formats is the Zarr format, which is designed to store large, chunked, compressed, N-dimensional arrays.
This repository provides Python scripts that convert the various raw data sets collected during ORCESTRA to Zarr.
This repository includes a uv project structure, which defines a strict set of dependencies. You can run each script by calling:
uv run <script>