API Reference¶

High-Level Functions¶

Function	Description
`secact_activity_inference()`	Bulk RNA-seq inference
`secact_activity_inference_scrnaseq()`	scRNA-seq inference
`secact_activity_inference_st()`	Spatial transcriptomics inference
`load_signature(name='secact')`	Load built-in signature matrix

Core Functions¶

Function	Description
`ridge()`	Single-call ridge regression with permutation testing
`ridge_batch()`	Batch processing for large datasets (dense or sparse)
`ridge_batch_streaming()`	Two-pass streaming inference for >5M-cell H5AD files
`H5ADChunkReader`	Memory-efficient chunk reader for H5AD files via h5py
`estimate_batch_size()`	Estimate optimal batch size for available memory
`estimate_memory()`	Estimate memory requirements

Key Parameters¶

Parameter	Default	Description
`sig_matrix`	`"secact"`	Signature: "secact", "cytosig", or DataFrame
`lambda_`	`5e5`	Ridge regularization parameter
`n_rand`	`1000`	Number of permutations
`seed`	`0`	Random seed for reproducibility
`rng_method`	`None`	RNG backend: `'srand'` (match R, default), `'gsl'` (cross-platform), `'numpy'` (fast). `None` defaults to srand.
`is_group_sig`	`True`	Group similar signatures by correlation before regression
`backend`	`'auto'`	'auto', 'numpy', or 'cupy'
`use_cache`	`False`	Cache permutation tables to disk
`sparse_mode`	`False`	Keep sparse Y in sparse format (avoids densification)
`col_center`	`True`	Subtract column means during sparse in-flight normalization
`col_scale`	`True`	Divide by column stds during sparse in-flight normalization

ST-Specific Parameters¶

Parameter	Default	Description
`cell_type_col`	`None`	Column in AnnData.obs for cell type
`is_spot_level`	`True`	If False, aggregate by cell type
`scale_factor`	`1e5`	Normalization scale factor

Batch Processing Parameters¶

Parameter	Default	Description
`batch_size`	`None`	Samples per batch (`None` = all at once)
`output_path`	`None`	Stream results to H5AD file (requires `batch_size`)
`output_compression`	`"gzip"`	Compression: "gzip", "lzf", or None

Streaming H5AD Parameters¶

These parameters are available on secact_activity_inference_scrnaseq() and secact_activity_inference_st() for processing >5M-cell datasets that exceed available RAM.

Parameter	Default	Description
`streaming`	`False`	Enable two-pass chunk reading from H5AD (requires file path input, not AnnData)
`streaming_chunk_size`	`50_000`	Number of cells per H5AD read chunk

For low-level ridge() / ridge_batch() usage and sparse column normalization details, see Advanced API.