Page 1 of 1

[SOLVED] How to perform a SGS (seq. gauss. simulation)

PostPosted: Mon Nov 23, 2015 7:41 pm
by Gustavo Ramos
Good afternoon.
I would like to know how to perform a sequential gaussian simulation in RGeostats.
I'm trying to compare the results of a ordinary kriging with the SGS but I can't find how to perform it within the RGeostats package.
Is there a way to do it?
Thanks for the attention.

Re: How to perform a SGS (sequential gaussian simulation)

PostPosted: Thu Nov 26, 2015 10:15 am
by Didier Renard
Sorry to say that there is no SGS simulation in the current version.
I imagine I will have to program it for the next version

Re: How to perform a SGS (sequential gaussian simulation)

PostPosted: Thu Nov 26, 2015 10:23 am
by Gustavo Ramos
Good morning.
Thanks for the help.

Re: [SOLVED] How to perform a SGS (seq. gauss. simulation)

PostPosted: Wed Jul 26, 2023 5:45 pm
by Jeffrey Yarus
Was SGS or SIS ever added to RGeostats?

Re: [SOLVED] How to perform a SGS (seq. gauss. simulation)

PostPosted: Thu Aug 03, 2023 5:55 pm
by Didier Renard
Hi Jeffrey
Sorry for this late answer. And sorry for this NEGATIVE answer.

Let us recall the main principle of the sequential (conditional) simulation... In turn, we consider a new target. We calculate the estimation at the target site, based on whatever information is available at the current time (i.e. the initial hard data + the values at the target sites processed before the current one). This estimation, to which we add a random term (obtained sampled from a normal distribution but tuned to the variance of the estimation error), produce the conditional simulation at the current target site.
This new simulated value will be part of the available information when processing the next target site.

In RGeostats (and still in future gstlearn), we have not been willing to add the Sequential simulation technique. Essentially for two reasons:
- a theoretical one: Unless we always consider ALL the previously simulated targets in the kriging procedure, the results of such a procedure is unsafe (unless we use specific models and pattern of information where screen effect is valid). Therefore the available information grows with the rank of the target site. This quickly leads to an issue as the kriging system becomes huge and inverting it becomes intractable. Then people usually introduce a problem reduction by introducing a *neighborhood*. The impact of this neighborhood on the quality of the result has not been seriouly analyzed (for my point of view).
- a practical one: at each sequence, the information comes on one hand from the initial data, and from the already simulated targets on the other hand. A proper implementation would require handling a double neighborhood ... which is not available in RGeostats (nor in gstlearn).

Sorry.

Re: [SOLVED] How to perform a SGS (seq. gauss. simulation)

PostPosted: Mon Aug 07, 2023 9:36 pm
by Jeffrey Yarus
Hi Didier: Don't apologize, I don't use SGS or SIS very much, principally for the reasons you state. I was just looking to see if you included it because some students were asking. That said, I continued to be surprised to find that so many people continue to use it...