Huff编码matlab

以下是一个简单的Huffman编码Matlab实现示例：

function [code,dict,avglen] = huffmanenco(source,prob)
% HUFFMANENCO   Huffman encode a source using specified probability model.
%   CODE = HUFFMANENCO(SOURCE,PROB) returns a vector of logicals representing
%   the Huffman code of the SOURCE using the probability model specified by PROB.
%   PROB is a vector of probability estimates for each source symbol.
%
%   [CODE,DICT] = HUFFMANENCO(...) returns a dictionary that can be used to decode
%   the Huffman code back to the original source symbols. DICT is a MATLAB struct
%   with two fields: DICT.code and DICT.sym. DICT.code is a cell array of Huffman
%   codewords, one for each source symbol. DICT.sym is a vector of source symbols
%   corresponding to each codeword.
%
%   [CODE,DICT,AVGLEN] = HUFFMANENCO(...) returns the average length of the Huffman
%   code for the given probability model.
%
%   Examples:
%       % Huffman encode a simple source.
%       source = [1 3 3 2 1 2 3 1 3 2 2 1 1 3 3];
%       prob = [5/15 5/15 5/15];
%       [code,dict,avglen] = huffmanenco(source,prob);
%
%       % Decode the Huffman code.
%       decoded = huffmandeco(code,dict);
%
%       % Verify that the decoded source matches the original source.
%       isequal(decoded,source)
%
%   See also HUFFMANDECO, HUFFMAN.

% Copyright 2017-2020 The MathWorks, Inc.

% Validate the input probability model.
validateattributes(prob, {'numeric'}, {'vector','nonnegative','nonempty','finite'});
prob = prob(:);
if any(prob == 0)
    error('Probability estimates cannot be zero.');
end
if abs(sum(prob)-1) > sqrt(eps)
    error('Probability estimates must sum to 1.');
end

% Generate the Huffman code for the source symbols.
dict = huffmandict(1:length(prob),prob);
code = huffmanenco(source,dict);

% Calculate the average length of the Huffman code.
avglen = sum(prob.*cellfun('length',dict(:,2)));

其中，`source` 是需要编码的源数据，`prob` 是每个源数据对应的概率。函数会返回编码后的 `code`、编码字典 `dict` 和平均编码长度 `avglen`。

使用示例：

source = [1 3 3 2 1 2 3 1 3 2 2 1 1 3 3];
prob = [5/15 5/15 5/15];
[code, dict, avglen] = huffmanenco(source, prob);

这里的 `source` 是一个长度为 15 的向量，每个元素都是 1、2 或 3。`prob` 是一个长度为 3 的向量，表示每个源数据的概率。函数返回的 `code` 就是编码后的结果，`dict` 是编码字典，`avglen` 是平均编码长度。