Communications Toolbox

dpcmopt

Optimize differential pulse code modulation parameters

Syntax

predictor = dpcmopt(trainingset,ord);
[predictor,codebook,partition] = dpcmopt(trainingset,ord,length);
[predictor,codebook,partition] =
dpcmopt(trainingset,ord,initcodebook);

Description

predictor = dpcmopt(trainingset,ord) returns a vector representing a predictive transfer function of order ord that is appropriate for the training data in the vector trainingset. predictor is a row vector of length ord+1. See Representing Quantization Parameters for more about its format.

Note    dpcmopt optimizes for the data in trainingset. For best results, trainingset should be similar to the data that you plan to quantize.

[predictor,codebook,partition] = dpcmopt(trainingset,ord,length) is the same as the syntax above, except that it also returns corresponding optimized codebook and partition vectors codebook and partition. length is an integer that prescribes the length of codebook. partition is a vector of length length-1. See either Representing Quantization Parameters or the reference page for quantiz, for a description of the formats of partition and codebook.

[predictor,codebook,partition] =
dpcmopt(trainingset,ord,initcodebook) is the same as the first syntax, except that it also returns corresponding optimized codebook and partition vectors codebook and partition. initcodebook, a vector of length at least 2, is the initial guess of the codebook values. The output codebook is a vector of the same length as initcodebook. The output partition is a vector whose length is one less than the length of codebook.

Examples

See the section Example: Comparing Optimized and Nonoptimized DPCM Parameters for an example that uses dpcmopt.

See Also
dpcmenco, dpcmdeco, quantiz, lloyds

dpcmenco

encode