Protein.run_lda

Protein.run_lda(layer, attribute, output_label, cycles=1)

Perform LDA on a row attribute using a layer to type clusters.

The row attribute is dimensionally reduced using UMAP.

Adds output_label to the row attributes.

Parameters

layerstr / np.array

The layer used for cluster typing. Uses _Assay.get_attribute() to retrieve the value constrained by row+col.

attributestr / np.array

The row attribute to be used for UMAP. Uses _Assay.get_attribute() to retrieve the value constrained by row.

output_labelstr

The name of the row attribute where the UMAP output is.

cyclesint

The number of rounds of LDA to perform.

kwargsdictionary

Passed to UMAP.