Lukas Heumos

@jlause do you plan on tackling these to get this in?

@Sakshi-2797 is there a reason why you'd introduce a flavor here instead of outright replacing the implementation?

Unfortunately, I run into ``` __________________________________________________________________________________ test_scale[use_fastpp] ___________________________________________________________________________________ flavor = 'use_fastpp' @pytest.mark.parametrize("flavor", ["default", "use_fastpp"]) def test_scale(flavor): adata = pbmc68k_reduced() adata.X = adata.raw.X v = adata[:, 0 : adata.shape[1] // 2]...

Please adapt the corresponding test to: ``` @pytest.mark.parametrize("flavor", ["default", "use_fastpp"]) def test_scale(flavor): adata = pbmc68k_reduced() adata.X = adata.raw.X v = adata[:, 0 : adata.shape[1] // 2] # Should turn view...

I pushed to your branch. It failed yesterday while Github was having issues. The test should pass.

@Sakshi-2797 were you able to find the fix to make it work or do you need more information? I'm happy to take over your branch as soon as it works...

Sparse matrices are absolutely essential for single-cell to ensure that RAM usage is kept low. Is there a way to make your implementation work with sparse matrices?

Could you please tell me how R is related to this? Are you attempting to use pertpy through reticulate?

Ah okay. So 1. You're not installing the latest pertpy version. Please run `pip install -U pertpy` 2. You're using Windows which is the root of all evil here. We...

Yeah, as mentioned above, we cannot provide Windows support at the moment. Please use WSL or a Linux machine for now.