Main -> Users' Guide and Examples
The software has been tested with MATLAB version 220.127.116.117 (R2009a). The software may not work with earlier releases that do not support object-oriented programming. Sorry, we are still looking for a workaround.
- Unzip the file
gampmatlab.zipin any directory.
- Add the subdirectory,
code/main, to your MATLAB path.
- In MATLAB,
estimTest. This will run the GAMP algorithm for a Bernoulli-Gaussian sparse vector and compare the performance against a linear least-squares estimator.
Examples of the code are found in the subdirectory
Right now, only two simple examples have been created, but we are in the process of putting in more
to illustrate various features of the code.
- Example 1: Getting started: Estimate a Gaussian vector. Illustrates the basic syntax of the
- Example 2: Estimating a sparse vector with AWGN measurements. Illustrates how to use the
SparseScaEstimfunction to estimate vectors with arbitrary sparse mixture distributions.
You can also take a look at the file
code/test/estimTest.m for an example of different estimators.