NVIDIA / NVIDIA/cudf

[FEA] polars collect_all(engine="gpu")

Open
#22,840 1 comment 0 reactions 0 assignees View on GitHub
bug cudf-polars feature request
Dominant language
C++
Stars
9.8k
Forks
1.1k
Avg merge
3d 6m
Merged PRs (30d)
278

Description

Polars `collect_all` [documents](https://docs.pola.rs/api/python/dev/reference/api/polars.collect_all.html) support for `engine="gpu"` but this doesn't actually work:
```py
import polars as pl

lf1 = pl.LazyFrame({"a": [1, 2, 3]}).filter(pl.col("a") > 1)
lf2 = pl.LazyFrame({"b": [10, 20, 30]}).select(pl.col("b") * 2)

pl.collect_all([lf1, lf2], engine="gpu")
```
yields
```
polars.exceptions.InvalidOperationError: collect_all is not supported for the gpu engine
```

This would be a nice feature to support; I imagine multiple collects could be accelerated well on the GPU.

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.