zellkonverter replaced with anndataR

[BilalsGituation]

First of all, great ebook!

I received a deprecation notice for zellkonverter when trying to run some code from this chapter and wanted to keep the chapter reproducible by "future-proofing" it a little with tools provided in anndataR.

I explained my reasoning for the edit in more detail in Issue#272 available at #272

I was able to test everything here, although my render had trouble reading the .h5ad file in chunk {python spatial-scatter}. I had success reading the file and saw what one would expect in there, using the anndataR package (edit: and anndata via reticulate using an absolute path), so I went ahead and committed without changing your version of the chunk.

There was another an error I had to suppress when running ad3 <- anndata$read_h5ad() in reticulate (chunk named r h5ad2Reticulate in my version), since it was mitigated by successful creation of the ad3 object.

Thank you for reading. Please feel free to comment either here or on Issue#272 if you have questions or feedback.

[HelenaLC]

Thank you for the suggestion and effort to draft a PR! I agree that replacing zellkonverter with anndataR is preferable, and this has been on our radar now that the package is in Bioc and stable.

We are currently facing issues with the book build related to the Python environment setup. I will look into the proposed PR after this has been resolved, as we currently cannot test the book build and deployment.

[BilalsGituation]

No problem, good luck finding a solution to the book build issue! I hope that my PR is useful once you can look into it.

[BilalsGituation]

Discovered and corrected a typo building the file spe.h5ad. I had written obs_mapping rather than obsm_mapping.
Retested file contents produced by chapter render using the following code, with anndata referring to the python module, exactly as called in the pre-PR chapter.

h5ad <- "inst/pages/spe.h5ad"
(adata <- anndata$read_h5ad(h5ad)) # ReticulateAnnData object with expected form + elements
adata$uns # shows image metadata for one sample
adata$obsm # shows spatial x and y coords for each spot
adata$obs # shows spot barcodes and columns "in_tissue array_row array_col sample_id library_id" in df-like format