← Back to context

Comment by Omega359

3 days ago

When I started on my path to get my application off of Spark late 2023 I started with Polars because it seemed to have more community velocity and seemed more approachable. Unfortunately, at least at that time, the lazy evaluation Rust api was very much a wip and didn't work for my use case. Switching to DataFusion enabled me to port/rewrite my application into Rust and drastically improve it's performance.