Struct FrequencyModel 

pub struct FrequencyModel {
    pub frequencies: Vec<u32>,
    pub cumulative: Vec<u32>,
    pub scale_bits: u8,
    pub active_symbols: usize,
}

A canonical sparse frequency model.

Fields

frequencies: Vec<u32>

Normalized frequencies, indexed by symbol.

cumulative: Vec<u32>

Cumulative frequencies (length 257).

scale_bits: u8

Scale bits used.

active_symbols: usize

Number of active (non-zero) symbols.

Implementations

impl FrequencyModel

pub fn build(histogram: &[u64; 256], scale_bits: u8) -> Result<Self, AppError>

Build a normalized model from a raw histogram.

Algorithm:

1. Count occurrences per symbol (0..255).
2. Compute quota: floor(count[sym] * total / total_count).
3. Enforce minimum 1 per observed symbol.
4. Distribute remainder in descending remainder order.
5. If over target due to minimum-1 enforcement, remove from symbols with smallest remainder, never going below 1.

This is fully deterministic — no floating point, no platform dependency.

pub fn to_bytes(&self) -> Vec<u8>

Serialize to canonical binary format.

pub fn from_bytes(bytes: &[u8], scale_bits: u8) -> Result<Self, AppError>

Parse from canonical binary format.

pub fn sha256(&self) -> [u8; 32]

Compute SHA-256 of the canonical serialization.

pub fn build_uniform(scale_bits: u8) -> Self

Build a uniform model: all 256 symbols get equal (or nearly equal) frequency.

Trait Implementations

impl Clone for FrequencyModel

fn clone(&self) -> FrequencyModel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FrequencyModel

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

Formats the value using the given formatter.

impl Eq for FrequencyModel

impl PartialEq for FrequencyModel

fn eq(&self, other: &FrequencyModel) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for FrequencyModel