Skip to main content

streaming_sparse_running_stats

Function streaming_sparse_running_stats 

Source
pub fn streaming_sparse_running_stats(
    data_vec: &SparseIoVec,
    block_size: Option<usize>,
    progress_label: &str,
) -> Result<SparseRunningStatistics<f32>>
Expand description

Stream cells through data_vec in block_size chunks, accumulating per-feature NB sufficient statistics in parallel and merging them in a fixed order ([fold_in_fixed_order]), so the result does not depend on the thread count.

progress_label is shown in the indicatif progress bar (e.g. "HVG", "NB-Fisher").