File Exchange

image thumbnail

Levinson-Durbin algorithm for vector processes

version 0.8.0 (4.45 KB) by somename
Calculation of optimal linear predictor matrices for stochastic vector processes.


Updated 12 Jul 2018

View License

Does what Matlab's lpc does, but for vector processes. Implements the Levinson-Durbin algorithm for vector processes and thus is a generalization of the lpc function. This submission includes three files. Quick and dirty implementation, bear with me. For an 1 x M vector (or an uncorrelated N x M matrix) this function returns identical coefficients to the lpc result.
Implementation according to "The theory of linear prediction", chapter 8 by P. P. Vaidyanathan.

- vec_lpc.m contains the actual algorithm to calculate the optimal linear vector prediction coefficients/matrices.
- AutoCorrVec.m contains the supplementary calculation of the Autocorrelation function of the vector process. Direct usage should not be needed.
- example_usage.m showcases how to use the vec_lpc.m function and compares the performance of the scalar prediction with the vectorial prediction. This comes with a matrix of testdata included in the file dummydata.mat, which exhibits strong crosscorrelation.

As far as I know this has not been submitted on here yet. If you have any questions or critique mail me to or comment on here. For usage see the comments in vec_lpc.m and example_usage.m.

Comments and Ratings (0)

MATLAB Release Compatibility
Created with R2018a
Compatible with any release
Platform Compatibility
Windows macOS Linux

Discover Live Editor

Create scripts with code, output, and formatted text in a single executable document.

Learn About Live Editor