Hello all,
I have a quick question that has been on my mind for a while : let's say that in my Python code, I have a large array which represents the raw results of a MCMC run. Then, after importing the GetDist module, I create a "MCSamples" object based on this array.
Now if I run these two lines of code :
> g = getdist.plots.getPlotter()
> g.triangle_plot(MC,names)
I was wondering if there was any kind of processing (for example some form of automatic thinning maybe) that was done on the chain by GetDist before it plots the corresponding histograms and 2D contours ?
Thanks in advance !
Stephane
[GetDist] Question about chain processing
-
- Posts: 17
- Joined: February 11 2014
- Affiliation: IRAP Toulouse
-
- Posts: 1944
- Joined: September 23 2004
- Affiliation: University of Sussex
- Contact:
Re: [GetDist] Question about chain processing
It doesn't do any thinning. It can remove burn it, but only if you have set the relevant setting.
(of course they aren't raw histograms, it does quite a bit of processing to try to get good kernel density estimates to plot)
(of course they aren't raw histograms, it does quite a bit of processing to try to get good kernel density estimates to plot)
-
- Posts: 17
- Joined: February 11 2014
- Affiliation: IRAP Toulouse
[GetDist] Question about chain processing
Thanks for the answer Antony !
That was precisely because I knew that the KDE method was pretty sophisticated, that I was wondering if it was doing something to the chains in the process...
That was precisely because I knew that the KDE method was pretty sophisticated, that I was wondering if it was doing something to the chains in the process...