dynamic-systems-and-gp / gpdyn Goto Github PK

lmgp part of GPdyn toolbox is tied to the standalone script "gpSD00.m". Find a way to make lmgp flexible to current gpml toolbox.

Exact propagation has an analytical solution if the integral of the kernel times a Gaussian function can be computed analytically. There is not a solution for unifying the calculation of exact propagation (in a single script) for such class of covariance functions.

More plausible solutions exist:

1. Each script is written on its own
2. There is a single script for propagation, but contains separated numerical recipes for various covariance functions
3. There is a single script for propagation that requires specific properties of propagation to be available inside each covariance function, for example, the integral of the kernel times a Gaussian function is already available inside the covariance function script.

Hi,

I am using this toolbox to fit dynamic gaussian processes and could see that every different run of trainGParx using "minimize.m" was generating a different set of means and standard deviations (a great discrepancy at each run). Then, I updated the folder "gmpl-matlab" with the newest version of GPML available (gpml-matlab-v4.2-2018-06-11) from http://www.gaussianprocess.org/ and not only the results of "minimize.m" are consistent at every run (With regard to the predicted mean and standard deviation) and it was way faster than using "minimizeDE.m".

The only thing that I had to do was to copy and past the "rewrap.m" function from the older GPML directory (the one provided in this repo) and paste it into the newest one.

Therefore, I suggest to update GPML folder with the newest one available.

the code about simulation with propagation of uncertainty "simulGPmcmc" is not Markov Chain Monte Carlo method, but only Monte Carlo.

some function handles loaded from .mat files contain information about irrelevant absolute paths. The function handles should be independent of the defined absolute path from the past usage on a local machine.