load_embedding

load_embedding(embedding_id, symbol_prefix='', num_pca_dims=None, hgnc_renaming=False)[source]

Load specified embedding as DataFrame.

NOTE: pl.DataFrame contain a column named ‘index’ that holds the symbols but it is not an actual index as polars DataFrames does not have the concept of an index.

Parameters:

  • embedding_id (str) – ID of the embedding.

  • symbol_prefix (str) – prefix to add on all symbols in the vocabulary.

  • num_pca_dims (Optional[int]) – If not None, reduce the embedding to this number of dimensions via PCA.

Return type:

DataFrame

Returns:

The corresponding embedding map.

Raises:

ValueError – when the given embedding identifier is not registered.