Contents

ALLCools.integration

ALLCools.integration

Package Contents

simple_cca(adata, group_col, n_components=50, random_state=0)[source]
incremental_cca(a, b, max_chunk_size=10000, random_state=0)[source]

Perform Incremental CCA by chunk dot product and IncrementalPCA

Parameters

  • a – dask.Array of dataset a

  • b – dask.Array of dataset b

  • max_chunk_size – Chunk size for Incremental fit and transform, the larger the better as long as MEM is enough

  • random_state –

Returns

Return type

Top CCA components

calculate_direct_confusion(left_part, right_part)[source]

Given 2 dataframe for left/source and right/target dataset, calculate the direct confusion matrix based on co-cluster labels. Each dataframe only contain 2 columns, first is original cluster, second is co-cluster. The returned confusion matrix will be the form of source-cluster by target-cluster.

Parameters

  • left_part –

  • right_part –

previous

ALLCools.gtf.utilities

next

ALLCools.integration.cca