Page 1 of 1

[SOLVED] How to generate N realisation of PGS

PostPosted: Mon Jul 19, 2021 11:51 am
by Miorat
Hello there,

Could anyone assist me on how to create 100 or N numbers of ensemble realisation Truncated Pluri Gaussian?

I have managed to create one PGS, but I would need 100 for my ensemble. Something like (100, 2500).

Thank you in advance,

Re: How to generate N number of ensemble realisation of PGS

PostPosted: Tue Jul 20, 2021 7:37 am
by Fabien Ors
Dear Miorat,
You should use nbsimu=100 or nbsimu=2500 in the simpgs function.
I suggest you to try a smaller number of simulations before launching for 100.
For more details, you can refer to the PGS vignette available here : http://rgeostats.free.fr/doc/PGS.html
Hope this helps

Re: How to generate N number of ensemble realisation of PGS

PostPosted: Tue Jul 20, 2021 10:00 am
by Miorat
Hi,

Thanks a lot for getting back to me. So, I tried to change the nbsimu for 10. But I ended up getting errors
Code: Select all
outgrid2 <- simpgs(dbout=outgrid2, rule=rule4fac, dbprop=outgrid2,
                   model1=modY1,model2=modY2, nbsimu=10,
                   nbtuba=600,seed=126543,verbose=TRUE)


Here is what I get, could you please indicate what is possibly the problem and how to go about it?

Code: Select all
The Space Dimension of the Db structure (2)
Does not correspond to the Space Dimension of the model (1457366288)

The function is interrupted
Traceback:  simpgs(dbout = outgrid2, rule = rule4fac, dbprop = outgrid2,      model1 = modY1, model2 = modY2, nbsimu = 10, nbtuba = 600,      seed = 126543, verbose = TRUE)
Error: Error in R_simpgs


Thanks again!

Re: How to generate N number of ensemble realisation of PGS

PostPosted: Wed Jul 21, 2021 1:15 pm
by Didier Renard
Hi
The information provided does not allow a correct understanding of the misfunctioning... We need to know more.
1) The argument 'dbprop' should be useless (as you do not invoke non-stationarity... see argument 'flag_stat').
Therefore can you run the script without this argument (but I really think that the problem should remain).
2) As you could see, the problem comes from the Models (no way to know which one in particular). More over, I am not certain that both Models are usefull (this depends upon the Rule that you have specified).
So please can you simply do:
print(modelY1)
print(modelY2)
beforehand... and add that to your post.
Thanks in advance.

Re: How to generate N number of ensemble realisation of PGS

PostPosted: Thu Jul 22, 2021 5:15 pm
by Miorat
Hi,

I have fixed the issue, can't really tell what was the problem previously but after running it several times it works now.
Thank you for taking time to reply!

Regards,