Problem using genmap with large numbers of elements
Dear Nek Users, We are about to set up a new case with about 29692 in 2D, which we would like to extrude to about 150 elements in 3D, giving a total of about 30k times 150 = 4.5M elements. n2to3 works fine, but genmap gives a "periodic mismatch error". Having a lower number of extrusion elements, say 30k times 70 = 2.1M elements, works fine. Changing the tolerance in genmap does not fix it. Before diving into the code of n2to3 and genmap, we wanted to ask whether somebody has experienced that before, and whether you'd know a potential fix for the problem. We used both an older version of genmap and the latest github version. Thanks a lot! Ricardo
Dear Ricardo, I believe there is a fix to this... I'm not 100% certain of the status in the repo. Perhaps one of the other developers is aware... Will try to find out more. Best, Paul ________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Wednesday, December 21, 2016 3:21 PM To: [email protected] Subject: [Nek5000-users] Problem using genmap with large numbers of elements Dear Nek Users, We are about to set up a new case with about 29692 in 2D, which we would like to extrude to about 150 elements in 3D, giving a total of about 30k times 150 = 4.5M elements. n2to3 works fine, but genmap gives a "periodic mismatch error". Having a lower number of extrusion elements, say 30k times 70 = 2.1M elements, works fine. Changing the tolerance in genmap does not fix it. Before diving into the code of n2to3 and genmap, we wanted to ask whether somebody has experienced that before, and whether you'd know a potential fix for the problem. We used both an older version of genmap and the latest github version. Thanks a lot! Ricardo _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
Yes, this bug is reported here https://github.com/Nek5000/Nek5000/issues/125 but has not been resolved yet. Stefan -----Original message-----
From:[email protected] <[email protected]> Sent: Thursday 22nd December 2016 0:23 To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Ricardo,
I believe there is a fix to this... I'm not 100% certain of the status in the repo.
Perhaps one of the other developers is aware...
Will try to find out more.
Best, Paul
________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Wednesday, December 21, 2016 3:21 PM To: [email protected] Subject: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Nek Users, We are about to set up a new case with about 29692 in 2D, which we would like to extrude to about 150 elements in 3D, giving a total of about 30k times 150 = 4.5M elements. n2to3 works fine, but genmap gives a "periodic mismatch error". Having a lower number of extrusion elements, say 30k times 70 = 2.1M elements, works fine. Changing the tolerance in genmap does not fix it.
Before diving into the code of n2to3 and genmap, we wanted to ask whether somebody has experienced that before, and whether you'd know a potential fix for the problem.
We used both an older version of genmap and the latest github version.
Thanks a lot! Ricardo
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
I usually run genmap in 64 bit precision --- To me, it looks like maketools does not build it this way ? (Or am I mis-reading the script?) I'm looking at tools/genmap/makefile ... genmap can be built with -r8 I'm not certain about the other tools. Certainly n2to3 could be build with -r8 Paul ________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Thursday, December 22, 2016 7:52 AM To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements Yes, this bug is reported here https://github.com/Nek5000/Nek5000/issues/125 but has not been resolved yet. Stefan -----Original message-----
From:[email protected] <[email protected]> Sent: Thursday 22nd December 2016 0:23 To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Ricardo,
I believe there is a fix to this... I'm not 100% certain of the status in the repo.
Perhaps one of the other developers is aware...
Will try to find out more.
Best, Paul
________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Wednesday, December 21, 2016 3:21 PM To: [email protected] Subject: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Nek Users, We are about to set up a new case with about 29692 in 2D, which we would like to extrude to about 150 elements in 3D, giving a total of about 30k times 150 = 4.5M elements. n2to3 works fine, but genmap gives a "periodic mismatch error". Having a lower number of extrusion elements, say 30k times 70 = 2.1M elements, works fine. Changing the tolerance in genmap does not fix it.
Before diving into the code of n2to3 and genmap, we wanted to ask whether somebody has experienced that before, and whether you'd know a potential fix for the problem.
We used both an older version of genmap and the latest github version.
Thanks a lot! Ricardo
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
OK... I see that genmap is indeed wdsize=8... never mind... ________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Thursday, December 22, 2016 2:32 PM To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements I usually run genmap in 64 bit precision --- To me, it looks like maketools does not build it this way ? (Or am I mis-reading the script?) I'm looking at tools/genmap/makefile ... genmap can be built with -r8 I'm not certain about the other tools. Certainly n2to3 could be build with -r8 Paul ________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Thursday, December 22, 2016 7:52 AM To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements Yes, this bug is reported here https://github.com/Nek5000/Nek5000/issues/125 but has not been resolved yet. Stefan -----Original message-----
From:[email protected] <[email protected]> Sent: Thursday 22nd December 2016 0:23 To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Ricardo,
I believe there is a fix to this... I'm not 100% certain of the status in the repo.
Perhaps one of the other developers is aware...
Will try to find out more.
Best, Paul
________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Wednesday, December 21, 2016 3:21 PM To: [email protected] Subject: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Nek Users, We are about to set up a new case with about 29692 in 2D, which we would like to extrude to about 150 elements in 3D, giving a total of about 30k times 150 = 4.5M elements. n2to3 works fine, but genmap gives a "periodic mismatch error". Having a lower number of extrusion elements, say 30k times 70 = 2.1M elements, works fine. Changing the tolerance in genmap does not fix it.
Before diving into the code of n2to3 and genmap, we wanted to ask whether somebody has experienced that before, and whether you'd know a potential fix for the problem.
We used both an older version of genmap and the latest github version.
Thanks a lot! Ricardo
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
I looked at the genmap issue, and I am quite sure that the fix is to simply comment out the lines c if(nelt.ge.1000000.and.cbl(f,e).eq.'P ') c $ call copyi4(bl(1,f,e),buf(5),1) !Integer ... as suggested also in the pdf on github, https://github.com/Nek5000/Nek5000/issues/125. The reason is that buf(5),buf(6) contain the element number as a real*8, so no conversion is needed. Therefore, all went ok up to about 2.5M elements, because then 4 bytes were sufficient to store the float. The corresponding lines for wdsizi=4 should remain, though. Philipp On 2016-12-22 21:36, [email protected] wrote:
OK... I see that genmap is indeed wdsize=8... never mind...
________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Thursday, December 22, 2016 2:32 PM To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements
I usually run genmap in 64 bit precision ---
To me, it looks like maketools does not build it this way ? (Or am I mis-reading the script?)
I'm looking at tools/genmap/makefile ...
genmap can be built with -r8
I'm not certain about the other tools. Certainly n2to3 could be build with -r8
Paul
________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Thursday, December 22, 2016 7:52 AM To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements
Yes, this bug is reported here https://github.com/Nek5000/Nek5000/issues/125 but has not been resolved yet. Stefan
-----Original message-----
From:[email protected] <[email protected]> Sent: Thursday 22nd December 2016 0:23 To: [email protected] Subject: Re: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Ricardo,
I believe there is a fix to this... I'm not 100% certain of the status in the repo.
Perhaps one of the other developers is aware...
Will try to find out more.
Best, Paul
________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Wednesday, December 21, 2016 3:21 PM To: [email protected] Subject: [Nek5000-users] Problem using genmap with large numbers of elements
Dear Nek Users, We are about to set up a new case with about 29692 in 2D, which we would like to extrude to about 150 elements in 3D, giving a total of about 30k times 150 = 4.5M elements. n2to3 works fine, but genmap gives a "periodic mismatch error". Having a lower number of extrusion elements, say 30k times 70 = 2.1M elements, works fine. Changing the tolerance in genmap does not fix it.
Before diving into the code of n2to3 and genmap, we wanted to ask whether somebody has experienced that before, and whether you'd know a potential fix for the problem.
We used both an older version of genmap and the latest github version.
Thanks a lot! Ricardo
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
participants (1)
-
nek5000-users＠lists.mcs.anl.gov