lstar

Convert / interchange single-cell & spatial omics data between formats with the lstar package — AnnData (.h5ad), Seurat (.rds), SingleCellExperiment, Conos, pagoda2, and .lstar.zarr stores — via the one-command `lstar convert` CLI or the in-language read_X/write_X profiles, bridged by lstar's uniform L* data model. Also covers heterogeneous-sample COLLECTIONS (Conos), what each conversion preserves vs. drops, version-robust format detection, and lazy/streamed per-gene stats.

Use when working with the lstar package or the L-star (L*) data model and Zarr interchange format for single-cell / spatial omics, ESPECIALLY to convert single-cell data between formats (AnnData / h5ad, Seurat, SingleCellExperiment, Conos, pagoda2) and languages (Python, R, C++). Covers converting/exporting/importing between those formats via profiles (read_anndata/write_anndata, read_seurat/write_seurat, read_sce/write_sce, write_conos/read_conos) or the one-command `lstar convert` CLI (with a fidelity report, native-acceptance check, and a package-free `--backend` fallback that converts .h5ad via h5py / Seurat & SCE .rds via base R, without anndata/SeuratObject/SingleCellExperiment), building datasets of axes and fields, assembling collections of heterogeneous samples from per-sample objects (collection_from) and converting a collection to a Seurat v5 split assay or a single AnnData, reading/writing .lstar.zarr stores, lazy/streaming reads, the C++ accelerator (libstar), per-gene reductions, and format/vers