Skip to contents

Handles the modern dataframe encoding (per-column categorical groups, nullable/masked integer & boolean groups, string and numeric arrays) as well as two legacy factor layouts (a shared __categories group and /uns/<col>_categories). The index dataset becomes the rownames.

Usage

quasar_h5_read_dataframe(filename, name, index_as_column = FALSE)

Arguments

filename

path to the .h5ad file.

name

group name, e.g. "obs", "var" or "raw/var".

index_as_column

keep the index as a column in addition to using it as rownames. Default FALSE (cleaner colData / meta.data).

Value

a data.frame with rownames taken from the anndata index.