Skip to content

fix: skip NaN scores in IVF top-k heaps #116

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
adrian-wang wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: skip NaN scores in IVF top-k heaps #116

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
71 changes: 67 additions & 4 deletions core/src/ivfflat.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -194,7 +194,7 @@ impl FlatTopKHeap {
}

fn push(&mut self, dist: f32, id: i64) {
if self.k == 0 {
if self.k == 0 || dist.is_nan() {
return;
}
if self.data.len() < self.k {
Expand All @@ -205,16 +205,16 @@ impl FlatTopKHeap {
.data
.iter()
.enumerate()
.max_by(|(_, a), (_, b)| a.0.partial_cmp(&b.0).unwrap())
.max_by(|(_, a), (_, b)| a.0.total_cmp(&b.0))
{
if dist < self.data[worst_idx].0 {
if dist.total_cmp(&self.data[worst_idx].0).is_lt() {
self.data[worst_idx] = (dist, id);
}
}
}

fn into_sorted(mut self) -> Vec<(f32, i64)> {
self.data.sort_by(|a, b| a.0.partial_cmp(&b.0).unwrap());
self.data.sort_by(|a, b| a.0.total_cmp(&b.0));
self.data
}
}
Expand Down Expand Up @@ -310,4 +310,67 @@ mod tests {
assert!(distances[0] > 0.0);
assert_eq!(distances[1], f32::MAX);
}

#[test]
fn test_flat_topk_heap_skips_both_nan_signs() {
let expected = [(0.25, 5), (0.5, 3), (1.0, 1), (2.0, 4)];
for k in [0, 1, 3, 5] {
let mut heap = FlatTopKHeap::new(k);
for (distance, id) in [
(f32::NAN, 10),
(-f32::NAN, 11),
(1.0, 1),
(f32::NAN, 12),
(0.5, 3),
(-f32::NAN, 13),
(2.0, 4),
(f32::NAN, 14),
(-f32::NAN, 15),
(0.25, 5),
] {
heap.push(distance, id);
}
assert_eq!(
heap.into_sorted(),
expected[..k.min(expected.len())],
"k={k}"
);
}
}

#[test]
fn test_ivfflat_inner_product_search_skips_both_nan_signs() {
for nan in [f32::NAN, -f32::NAN] {
for (data, ids) in [
([nan, 1.0, 2.0], [99, 1, 2]),
([1.0, nan, 2.0], [1, 99, 2]),
([1.0, 2.0, nan], [1, 2, 99]),
] {
let mut index = IVFFlatIndex::new(1, 1, MetricType::InnerProduct);
index.set_quantizer_centroids(vec![0.0]);
index.add(&data, &ids, 3);
for k in 1..=3 {
let mut distances = vec![0.0; k];
let mut labels = vec![0; k];
index.search(&[1.0], 1, k, 1, &mut distances, &mut labels);
assert_eq!(labels, [2, 1, -1][..k]);
assert_eq!(distances, [-2.0, -1.0, f32::MAX][..k]);
}
}
}
}

#[test]
fn test_ivfflat_search_with_only_nan_returns_missing_neighbors() {
for metric in [MetricType::L2, MetricType::InnerProduct] {
let mut index = IVFFlatIndex::new(1, 1, metric);
index.set_quantizer_centroids(vec![0.0]);
index.add(&[f32::NAN, -f32::NAN], &[1, 2], 2);
let mut distances = [0.0; 3];
let mut labels = [0; 3];
index.search(&[1.0], 1, 3, 1, &mut distances, &mut labels);
assert_eq!(labels, [-1; 3]);
assert_eq!(distances, [f32::MAX; 3]);
}
}
}
51 changes: 47 additions & 4 deletions core/src/ivfflat_io.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1410,7 +1410,7 @@ impl ReaderTopKHeap {

#[inline]
fn push(&mut self, dist: f32, id: i64) {
if self.k == 0 {
if self.k == 0 || dist.is_nan() {
return;
}
if self.data.len() < self.k {
Expand All @@ -1423,7 +1423,7 @@ impl ReaderTopKHeap {
let worst_index = self
.worst_index
.expect("a full IVF-FLAT top-k heap has a worst entry");
if dist < self.data[worst_index].0 {
if dist.total_cmp(&self.data[worst_index].0).is_lt() {
self.data[worst_index] = (dist, id);
self.refresh_worst();
}
Expand All @@ -1434,12 +1434,12 @@ impl ReaderTopKHeap {
.data
.iter()
.enumerate()
.max_by(|(_, a), (_, b)| a.0.partial_cmp(&b.0).unwrap())
.max_by(|(_, a), (_, b)| a.0.total_cmp(&b.0))
.map(|(index, _)| index);
}

fn into_sorted(mut self) -> Vec<(f32, i64)> {
self.data.sort_by(|a, b| a.0.partial_cmp(&b.0).unwrap());
self.data.sort_by(|a, b| a.0.total_cmp(&b.0));
self.data
}
}
Expand Down Expand Up @@ -2396,4 +2396,47 @@ mod tests {
};
assert!(err.to_string().contains("reserved bytes must be zero"));
}

#[test]
fn test_reader_topk_heap_skips_both_nan_signs() {
let expected = [(0.25, 5), (0.5, 3), (1.0, 1), (2.0, 4)];
for k in [0, 1, 3, 5] {
let mut heap = ReaderTopKHeap::new(k);
for (distance, id) in [
(f32::NAN, 10),
(-f32::NAN, 11),
(1.0, 1),
(f32::NAN, 12),
(0.5, 3),
(-f32::NAN, 13),
(2.0, 4),
(f32::NAN, 14),
(-f32::NAN, 15),
(0.25, 5),
] {
heap.push(distance, id);
}
assert_eq!(
heap.into_sorted(),
expected[..k.min(expected.len())],
"k={k}"
);
}
}

#[test]
fn test_ivfflat_reader_inner_product_search_skips_both_nan_signs() {
for nan in [f32::NAN, -f32::NAN] {
let mut index = IVFFlatIndex::new(1, 1, MetricType::InnerProduct);
index.set_quantizer_centroids(vec![0.0]);
index.add(&[1.0, nan, 2.0], &[1, 2, 3], 3);
let mut reader =
IVFFlatIndexReader::open(Cursor::new(serialized_flat_index(&index))).unwrap();
for k in 1..=3 {
let (labels, distances) = reader.search(&[1.0], k, 1).unwrap();
assert_eq!(labels, [3, 1, -1][..k]);
assert_eq!(distances, [-2.0, -1.0, f32::MAX][..k]);
}
}
}
}
37 changes: 32 additions & 5 deletions core/src/ivfpq.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3322,7 +3322,7 @@ impl TopKHeap {

#[inline]
fn push(&mut self, dist: f32, id: i64) {
if self.k == 0 {
if self.k == 0 || dist.is_nan() {
return;
}
if self.data.len() < self.k {
Expand All @@ -3331,14 +3331,14 @@ impl TopKHeap {
build_max_heap(&mut self.data);
self.built = true;
}
} else if dist < self.data[0].0 {
} else if dist.total_cmp(&self.data[0].0).is_lt() {
self.data[0] = (dist, id);
sift_down(&mut self.data, 0);
}
}

fn into_sorted(mut self) -> Vec<(f32, i64)> {
self.data.sort_by(|a, b| a.0.partial_cmp(&b.0).unwrap());
self.data.sort_by(|a, b| a.0.total_cmp(&b.0));
self.data
}
}
Expand Down Expand Up @@ -3419,10 +3419,10 @@ fn sift_down(heap: &mut [(f32, i64)], mut i: usize) {
let left = 2 * i + 1;
let right = 2 * i + 2;

if left < n && heap[left].0 > heap[largest].0 {
if left < n && heap[left].0.total_cmp(&heap[largest].0).is_gt() {
largest = left;
}
if right < n && heap[right].0 > heap[largest].0 {
if right < n && heap[right].0.total_cmp(&heap[largest].0).is_gt() {
largest = right;
}
if largest == i {
Expand Down Expand Up @@ -6246,4 +6246,31 @@ mod tests {
fn ivfpq_rejects_odd_4bit_subquantizer_count_at_construction() {
let _ = IVFPQIndex::with_nbits(12, 4, 3, 4, MetricType::L2, false);
}

#[test]
fn test_topk_heap_skips_both_nan_signs() {
let expected = [(0.25, 5), (0.5, 3), (1.0, 1), (2.0, 4)];
for k in [0, 1, 3, 5] {
let mut heap = TopKHeap::new(k);
for (distance, id) in [
(f32::NAN, 10),
(-f32::NAN, 11),
(1.0, 1),
(f32::NAN, 12),
(0.5, 3),
(-f32::NAN, 13),
(2.0, 4),
(f32::NAN, 14),
(-f32::NAN, 15),
(0.25, 5),
] {
heap.push(distance, id);
}
assert_eq!(
heap.into_sorted(),
expected[..k.min(expected.len())],
"k={k}"
);
}
}
}
53 changes: 49 additions & 4 deletions core/src/topk.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,11 +46,11 @@ impl TopKHeap {
}

pub(crate) fn push(&mut self, dist: f32, id: i64) {
if self.k == 0 || dist >= self.max_distance {
if self.k == 0 || dist.is_nan() || dist >= self.max_distance {
return;
}
if let Some(&idx) = self.positions.get(&id) {
if dist < self.data[idx].0 {
if dist.total_cmp(&self.data[idx].0).is_lt() {
self.data[idx].0 = dist;
self.sift_down(idx);
}
Expand All @@ -62,7 +62,7 @@ impl TopKHeap {
self.sift_up(self.data.len() - 1);
return;
}
if dist < self.data[0].0 {
if dist.total_cmp(&self.data[0].0).is_lt() {
let old_id = self.data[0].1;
self.positions.remove(&old_id);
self.data[0] = (dist, id);
Expand All @@ -73,10 +73,11 @@ impl TopKHeap {

pub(crate) fn should_consider(&self, lower_bound: f32) -> bool {
self.k > 0
&& !lower_bound.is_nan()
&& if self.data.len() < self.k {
lower_bound < self.max_distance
} else {
lower_bound < self.data[0].0
lower_bound.total_cmp(&self.data[0].0).is_lt()
}
}

Expand Down Expand Up @@ -190,4 +191,48 @@ mod tests {

assert_eq!(heap.into_sorted(), vec![(3.0, 3), (4.0, 2)]);
}

#[test]
fn test_topk_heap_skips_both_nan_signs() {
let expected = [(0.25, 5), (0.5, 3), (1.0, 1), (2.0, 4)];
for k in [0, 1, 3, 5] {
let mut heap = TopKHeap::new(k);
for (distance, id) in [
(f32::NAN, 10),
(-f32::NAN, 11),
(1.0, 1),
(f32::NAN, 12),
(0.5, 3),
(-f32::NAN, 13),
(2.0, 4),
(f32::NAN, 14),
(-f32::NAN, 15),
(0.25, 5),
] {
heap.push(distance, id);
}
assert_eq!(
heap.into_sorted(),
expected[..k.min(expected.len())],
"k={k}"
);
}
}

#[test]
fn test_topk_heap_rejects_nan_for_duplicates_and_collector() {
for nan in [f32::NAN, -f32::NAN] {
let mut heap = TopKHeap::new(2);
assert!(!heap.should_consider(nan));
heap.push(6.0, 2);
assert!(!heap.should_consider(nan));
heap.push(8.0, 1);
assert!(!heap.should_consider(nan));
heap.push(nan, 2);
Collector::push(&mut heap, 3, nan).unwrap();
heap.push(4.0, 1);
heap.push(5.0, 3);
assert_eq!(heap.into_sorted(), vec![(4.0, 1), (5.0, 3)]);
}
}
}