Struct LearnedSparseIndex 

pub struct LearnedSparseIndex { /* private fields */ }

Learned Sparse Index using linear regression

Now with key normalization for numerical stability with large u64 keys. Keys are normalized to [0, n-1] before regression, avoiding precision loss with values near u64::MAX.

Implementations

impl LearnedSparseIndex

pub fn empty() -> Self

Create an empty index

pub fn build(keys: &[u64]) -> Self

Build index from sorted keys

Keys MUST be sorted in ascending order. This is not validated for performance.

pub fn build_with_threshold(keys: &[u64], correction_threshold: usize) -> Self

Build index with custom correction threshold

Lower threshold = more corrections stored = faster lookups but more memory Higher threshold = fewer corrections = slower lookups but less memory

pub fn lookup(&self, key: u64) -> LookupResult

Lookup: O(1) expected, O(ε) worst case

pub fn lookup_with_error(&self, key: u64, max_error: usize) -> LookupResult

Lookup with a custom error margin (tighter range if you know data is regular)

pub fn stats(&self) -> LearnedIndexStats

Get index statistics

pub fn is_efficient(&self) -> bool

Returns true if this index is well-suited for learned indexing (low error, few corrections)

pub fn memory_bytes(&self) -> usize

Memory usage in bytes

pub fn insert(&mut self, key: u64, position: usize, keys: &[u64]) -> bool

Incrementally add a key (rebuilds if model quality degrades) Returns true if a rebuild was triggered

Trait Implementations

impl Clone for LearnedSparseIndex

fn clone(&self) -> LearnedSparseIndex

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for LearnedSparseIndex

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for LearnedSparseIndex

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for LearnedSparseIndex

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.