Function super_mass::mass_batch

pub fn mass_batch<T: MassType>(
    ts: &[T], 
    query: &[T], 
    batch_size: usize, 
    top_matches: usize
) -> Vec<(usize, f64)>

Masss batch finds top subsequence per batch the lowest distance profile for a given query and returns the top K subsequences. This behavior is useful when you want to filter adjacent suboptimal subsequences in each batch, where the local optimum overlaps with suboptima differing only by a few index strides. This method implements MASS V3 where chunks are split in powers of two and computed in parallel. Results are partitioned and not sorted, you can sort them afterwards if needed.