ndjson
- Dominant language
- Jupyter Notebook
- Stars
- 2.6k
- Forks
- 213
- PR merge metrics
- No merged PRs in 30d
Description
Implement ndjson https://docs.pola.rs/api/python/stable/reference/api/polars.read_ndjson.html and https://docs.pola.rs/api/python/stable/reference/api/polars.DataFrame.write_ndjson.html as a data saver/loader.
It should go into https://github.com/DAGWorks-Inc/hamilton/blob/main/hamilton/plugins/polars_post_1_0_0_extensions.py, one class for the writer, on class for the reader, e.g. much like how the [polars json reader and writer ](https://github.com/DAGWorks-Inc/hamilton/blob/main/hamilton/plugins/polars_post_1_0_0_extensions.py#L487-L541) are structured.
There should be requisite tests to exercise the functionality. We can do one test that writes & then reads it's own output.
Then an example should be added to the directory called materialization under [example/polars](https://github.com/DAGWorks-Inc/hamilton/tree/main/examples/polars/). It should mirror the current structure in there.
If there's an issue with type hints, let me know and we can chat through it.
Contributor guide
Assessment
This issue has not been assessed yet.