Method build_pq_for_index

nodedb-vector/src/collection/quantize.rs:86–108

Train a PQ codec from a built HNSW index's live vectors.

(index: &HnswIndex, pq_m: usize)

    /// Train a PQ codec from a built HNSW index's live vectors.
    pub fn build_pq_for_index(index: &HnswIndex, pq_m: usize) -> Option<(PqCodec, Vec<u8>)> {
        let dim = index.dim();
        if pq_m == 0 || !dim.is_multiple_of(pq_m) {
            return None;
        }
        let n = index.len();
        let mut refs: Vec<Vec<f32>> = Vec::with_capacity(n);
        for i in 0..n {
            if !index.is_deleted(i as u32)
                && let Some(v) = index.get_vector(i as u32)
            {
                refs.push(v.to_vec());
            }
        }
        if refs.is_empty() {
            return None;
        }
        let refs_slices: Vec<&[f32]> = refs.iter().map(|v| v.as_slice()).collect();
        let k = 256usize.min(refs.len());
        let codec = PqCodec::train(&refs_slices, dim, pq_m, k, 20);
        let codes = codec.encode_batch(&refs_slices).ok()?;
        Some((codec, codes))
    }
}

// Keep the DEFAULT_SEAL_THRESHOLD import live when future refactors move