Implement higher level api to allow benchmarking multiple embeddings of the same dataset
Closed this issue · 1 comments
adamgayoso commented
Implement higher level api to allow benchmarking multiple embeddings of the same dataset
adamgayoso commented
Right now I'm imagining this would take an anndata and a list of obsm keys would denote several embeddings of the same data.
class Benchmarker:
def __init__(
adata:AnnData,
embedding_obsm_keys: Sequence[str],
bio_conservation_metrics: DataClass,
):
self._results = pd.DataFrame()
....