Installation#

zcollection 2.x is a Zarr v3-native rewrite of the library. It targets modern Python and the modern scientific stack; there is no Zarr v2 compatibility runtime.

Requirements#

Python 3.14 or later
zarr ≥ 3.1.6
numpy ≥ 1.20
numcodecs ≥ 0.13
xarray ≥ 2024.10
pandas
dask ≥ 2022.8 and distributed
fsspec

Optional extras#

pyarrow — required by the Parquet-backed Indexer.
obstore ≥ 0.5 — recommended S3 backend (s3:// URLs). Install with:
```
pip install "zcollection[obstore]"
```
icechunk ≥ 2.0 — transactional, multi-writer store (icechunk:// URLs):
```
pip install "zcollection[icechunk]"
```
s3fs — alternative S3 backend through fsspec, used by the test suite when running against a local minio server.

Install from PyPI#

pip install zcollection

To pull in the optional backends at the same time:

pip install "zcollection[obstore,icechunk]"

Install from conda-forge#

conda install -c conda-forge zcollection

Install from source#

Clone the repository and install in editable mode:

git clone https://github.com/CNES/zcollection.git
cd zcollection
pip install -e ".[test]"

The test extra adds pytest, pytest-cov and pytest-asyncio.

Running the tests#

From the repository root:

pytest zcollection/tests -q

No project-specific pytest options are required. Cloud-store tests are opt-in and described in their own modules.

Building the documentation#

The docs use Sphinx with Google-style docstrings. From the docs directory:

make html

The rendered HTML lands in docs/build/html.

Installation

Contents