8000 Replacing pandas with polars · Issue #1926 · pvlib/pvlib-python · GitHub
[go: up one dir, main page]

Skip to content
Replacing pandas with polars #1926
Closed
Closed
@langestefan

Description

@langestefan

Is your feature request related to a problem? Please describe.
Pandas is used extensively in pvlib all over the place. However after using it for many years I think it is now outdated and needs to be replaced. It is greedy, slow, inconsistent and the API is awkward and difficult to use.

Describe the solution you'd like
I would like to propose replacing pandas with polars. Polars is a DataFrame interface built in Rust with many, many performance benefits. Too many to list them here, so I suggest visiting their page to learn more.

I know this is a very big change, but I believe it will lead to some serious performance and usability benefits. So there is no real plan yet, I would like to just start the discussion and see what others think. I think a good starting point would be to benchmark some performance-critical pvlib code in pandas and polars so we can quantify what the performance benefits for pvlib would be.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions

      0