Struct TopKHeap 

pub struct TopKHeap<T, F>
where
    F: Fn(&T, &T) -> Ordering,
{ /* private fields */ }

A bounded heap for streaming top-K selection

This maintains the K smallest (or largest) elements seen so far, without storing all N elements.

Complexity

• Push: O(log K) when heap is full, O(log K) insertion
• Drain: O(K log K) to produce sorted output
• Space: O(K)

Implementations

impl<T, F> TopKHeap<T, F>

where F: Fn(&T, &T) -> Ordering,

pub fn new(k: usize, comparator: F, want_smallest: bool) -> Self

Create a new top-K heap

• k: Number of elements to keep
• comparator: Defines the desired output order (Less = should come first)
• want_smallest: If true, keep the K elements that compare as smallest

pub fn push(&mut self, value: T)

Push an element into the heap

Complexity: O(log K)

pub fn threshold(&self) -> Option<&T>

Get the current boundary value

This is the value that new elements must beat to be included.

pub fn is_full(&self) -> bool

Check if the heap is at capacity

pub fn into_sorted_vec(self) -> Vec<T>

Drain the heap into a sorted vector

Complexity: O(K log K)

pub fn len(&self) -> usize

Current number of elements

pub fn is_empty(&self) -> bool

Check if empty