Different results on two computers

40 views
Skip to first unread message

Dominique Soudant

unread,
Jul 23, 2025, 6:15:09 AMJul 23
to R-inla discussion group
Hi

Capture d’écran du 2025-07-23 12-07-15.png
On the left, the result on a 13-year-old HP Z620 workstation, on the right on a dell precision 3581 acquired this year.  Note the 0 standard deviations on the left.

Both machines are on ubuntu 22.04. I'll collect the other information (e.g. R version, R-INLA version, processors) and indicate them later.
DS

Havard Rue

unread,
Jul 23, 2025, 7:35:30 AMJul 23
to R-inla discussion group, Dominique Soudant
The zero sd is certainly an error.  Are both on R-4.5 and the most recent testing version?

-- 
Håvard Rue 
On 23 Jul 2025 at 12:15 +0200, Dominique Soudant <dominique.so...@gmail.com>, wrote:
Hi
--
You received this message because you are subscribed to the Google Groups "R-inla discussion group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to r-inla-discussion...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/r-inla-discussion-group/85c58ac1-6340-49cf-9e4f-142a7e1ee1e9n%40googlegroups.com.

Finn Lindgren

unread,
Jul 23, 2025, 7:35:44 AMJul 23
to Dominique Soudant, R-inla discussion group
The INLA and R versions would be the most informative; since all the other columns are the same (up to very minor numerical precision differences) this looks like a temporary bug in an older inla version, as it seems to have computed everything correctly but just stored zeros for the sd in the output.
Finn

On 23 Jul 2025, at 12:15, Dominique Soudant <dominique.so...@gmail.com> wrote:


Hi
--
You received this message because you are subscribed to the Google Groups "R-inla discussion group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to r-inla-discussion...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/r-inla-discussion-group/85c58ac1-6340-49cf-9e4f-142a7e1ee1e9n%40googlegroups.com.

Dominique Soudant

unread,
Jul 23, 2025, 8:34:07 AMJul 23
to R-inla discussion group
HI, 

HP Z620
Processeur : Intel(R) Xeon(R) CPU E5-2620 0 @ 2.00GHz
Mémoire : 65859MB (1551MB utilisé)
R version 4.5.1 (2025-06-13)
This is INLA_24.12.11 built 2024-12-11 19:58:26 UTC.

This is the binary version of INLA that is installed here. I did this following advice from Havard:


Dell Precision
Processeur : 13th Gen Intel(R) Core(TM) i7-13800H
Mémoire : 32541MB (3859MB utilisé)
R version 4.5.1 (2025-06-13)
This is INLA_24.12.11 built 2024-12-11 19:58:26 UTC.

Here, INLA has been compiled.

Thanks in advance for your help. 
DS

Dominique Soudant

unread,
Jul 24, 2025, 3:56:22 AMJul 24
to R-inla discussion group
Hi, 
I ran 
inla.upgrade()
=> glibC issue
I ran
inla.binary.install() 
no more glibC issue and get my SDs back. 

Thank you for all your job !
DS
Reply all
Reply to author
Forward
0 new messages