Bag Generic Typing

Dask Bag
1

How would people feel about making dask.bag.Bag a generic type, so that type checkers and function signatures could include the information about what’s in a dask bag. For example

def add_lengths(bag: dask.bag.Bag[list[dict[str, Any]]]) -> dask.bag.Bag[list[dict[str, Any]]]:
    return bag.map(add_length)

def add_length(item: dict[str, Any]) -> dict[str, Any]:
    item["length"] = len(item["document"])

I’d be happy to write a PR for this.

2

Hi @jordan-jack-schneide, welcome to Dask community!

Thanks for proposing this!

As this is a technical proposal for Dask new feature, I would recommend you raise directly an issue on github for this, you might get faster and direct feedback from core developers.

3

Ah thanks! Wasn’t sure what the right place was. I’ll copy this over there.

1 Like