ViewmastR inference — viewmastR

This function performs cell type inference using a trained model, passing sparse matrices directly to Rust for maximum performance.

Usage

viewmastR_infer(
  query_cds,
  model_dir,
  selected_features,
  query_celldata_col = "viewmastR_inferred",
  labels = NULL,
  verbose = TRUE,
  return_probs = FALSE,
  return_type = c("object", "list"),
  batch_size = NULL,
  threads = 1
)

Arguments

query_cds: Seurat or cell_data_set object
model_dir: Path to the trained model directory
selected_features: Character vector of feature names
query_celldata_col: Name of column to store results (default: "viewmastR_inferred")
labels: Optional character vector of class labels
verbose: Print progress messages
return_probs: If TRUE, add probability columns to metadata
return_type: "object" returns the modified object, "list" returns object and probs
batch_size: Cells per inference batch (default: auto)
threads: Number of parallel threads (default: 1)

Value

Modified query_cds with inferred labels