Users

users@lists.einsteintoolkit.org

3 participants
3983 discussions

Einstein Toolkit Meeting Reminder
by users＠einsteintoolkit.org 29 Jul '26

29 Jul '26

Hello, Please consider joining the weekly Einstein Toolkit phone call at 9:00 am US central time on Thursdays. For details on how to connect and what agenda items are to be discussed, use the link below. https://docs.einsteintoolkit.org/et-docs/Main_Page#Weekly_Users_Call --The Maintainers

1 0

Agenda for Thursday's Meeting
by users＠einsteintoolkit.org 27 Jul '26

27 Jul '26

Please update the Wiki with agenda items for Thursday's meeting. Thanks! https://docs.einsteintoolkit.org/et-docs/meeting_agenda --The Maintainers

1 0

Issues configuring SimFactory for PBS cluster (Hubble) - Jobs
by MAKINO Taiki 27 Jul '26

27 Jul '26

Dear Einstein Toolkit Users, I previously sent the e-mail about the issues configuring SimFactory for PBS cluster a few days ago. In case it was overlooked or got lost, I am re-sending my message below. I have prepared hubble.ini, .cfg, .sub, .run files and edited some lines in hubble.ini and hubble.sub based on your advice. (Hubble is my lab’s cluster.) But some problems appear. .out files exist under the Cactus directory and they are named ‘SubmitScript.o123456’ (where 123456 is the PBS job_id). Also, when I checked my_simulation/SIMFACTORY/run/SubmitScript file inside the simulation output, I found that the its content was same as generic.sub. Why is SimFactory ignoring my change to the submitscript option and still applying the generic script? Is there a command I need to run to force SimFactory to reload the updated .ini configuration? Additionally, is there a documentation webpage that explains the variables enclosed in @ (like @JOB_ID@ or @SCRIPTFILE@)? I would like to know the full list of available placeholders. I have attached the ‘SubmitScript' file and ‘generic.sub' file for your reference. Thank you for your help. Best regards, Taiki Makino Graduate School of Science, Nagoya University makino.taiki.g6(a)s.mail.nagoya-u.ac.jp<mailto:makino.taiki.g6@s.mail.nagoya-u.ac.jp> 取得 Outlook for Mac<https://aka.ms/GetOutlookForMac>

1 1

Question about how to generate initial data with an arbitrary EOS
by Beyhan Karakaş 26 Jul '26

26 Jul '26

Hello Taiki, Here are the resources to create BNS initial data. Einstein Toolkit supports initial data generated with Lorene, SGRID and Fuka. Here is the Joshua Faber’s files for BNS ID: https://ccrgpages.rit.edu/~jfaber/BNSID Here are the recordings on BNS initial data with Lorene: - https://www.youtube.com/watch?v=3euS5yMBFmI - https://www.youtube.com/watch?v=Lq50IlXBJf8 - https://www.youtube.com/watch?v=-NUD-piBeek Here is the Fuka tutorial for generating BNS initial data: https://samueltootle.github.io/fuka/bnsxcts.html Best regards, Beyhan. Dear Einstein Toolkit Users, > > I am a new Einstein Toolkit user, and I would like to ask how to generate > initial data with an arbitrary neutron star EOS. > > My purpose is to perform binary neutron star simulations with a tabulated > or an analytical EOS(for example, a BSk EOS or another tabulated EOS that I > prepare myself). > > Could you tell me the recommended way to generate such initial data for > use with Einstein Toolkit? If there is a related document or examples, I > would appreciate it if you could tell me. > > Thank you very much for your time and help. > > Best regards, > > > Taiki Makino > > Graduate School of Science, Nagoya University > > makino.taiki.g6(a)s.mail.nagoya-u.ac.jp<mailto: > makino.taiki.g6(a)s.mail.nagoya-u.ac.jp> > > >

2 1

Question about IllinoisGRMHD C2P failure and kuibit Psi4 extraction
by MAKINO Taiki 23 Jul '26

23 Jul '26

Dear Einstein Toolkit Users, I am a new Einstein Toolkit user, and I have some questions about a BNS simulation. I am trying to run a BNS simulation using bns.par, which I downloaded from Einstein Toolkit Gallery (https://einsteintoolkit.org/gallery/bns/index.html) and slightly modified. I can successfully create and submit the simulation but the following message appears during the simulation: INFO (IllinoisGRMHD): C2P: Iter. # 0, Lev: 5 NumPts= 132300 | Backups: 0 0 0 | Fixes: VL= 0 rho*= 0 Averaged pts = 38250 Font1D 38250 | Failures: 2250 InHoriz= 0 / 0 | 1.00 iters/gridpt Error, Sum: rho 6.378e-08, 2.433e-05 | tau nan, -nan | Sx nan, nan | Sy nan, nan | Sz nan, nan To investigate this problem, I added the following parameters: Multipole::interpolator_name = "Lagrange polynomial interpolation" Multipole::interpolator_pars = "order=4" I also changed the parameter Driver::ghost_size from 3 to 4, but the same error occurred. I would like to ask: * What could cause tau, Sx, Sy, and Sz to become NaN in this situation? * Does this indicate a failure of C2P conversion? * Are there any parameter settings or initial data that I should check first? Additionally, I have another question. When I use kuibit to extract Psi4, I expected to obtain the full time series (or all available modes) but the returned object contains only one element, for example, SimDir.gws[100][(2,2)].y=[-9.79586237e-15-5.27476073e-14j]. Other attributes such as SimDir.gws[100][(2,2)].x also contain only one element. Is this a correct behavior or could it indicate a problem with my .par file? For reference, I have attached my parameter file. Thank you very much for your time and any suggestions. Best regards, Taiki Makino Graduate School of Science, Nagoya University makino.taiki.g6(a)s.mail.nagoya-u.ac.jp<mailto:makino.taiki.g6@s.mail.nagoya-u.ac.jp> 取得 Outlook for Mac<https://aka.ms/GetOutlookForMac> 取得 Outlook for Mac<https://aka.ms/GetOutlookForMac>

3 5

Meeting minutes for 07/23/26
by Lucas Timotheo Sanches 23 Jul '26

23 Jul '26

Attendees: Lucas (minutes), Roland (chair), Beyhan, Maxwell Rizzo, Peter, Deborah # Accumulated updates after the release - CarpetX (and AMReX updates): Liwei found that you can make almost all the changes so that they are backwards compatible. He submitted a PR with a patch demonstrating this. We will test this in a number of architectures to make sure it all works. - McLachlan updates: Peter had something he was testing. Then his laptop died. He has not come around to that yet, but he will come back to it. # Atlassian retiring Bitbucket Cloud Issues - Zach says: He did a wrangle of clankers to help him migrate things over. He completed that about 4 days ago. He's working on a private repo for the ETK. - What happens when user names don't match between systems? Zach did a public web search for publicly available data to find the correct user names. - Subtlety: As soon this is moved to a public repo, everyone that has a ticket, either open or closed, will get a notification for *each ticket*. That would be a catastrophe to people like Roland and Steve. - Who exactly is assigned to the bug? There's no way that Zach can make it as if user X opened a ticket. Zach will be listed as the ticket opener. The solution to this is to add text in the start of the ticket, saying it is ported from bitbucket, who was the original reporter, link to GH user page and other info, like original ticket number, ticket creation metadata and such. - He proposed to use a bot, or an artificial user that is the responsible for the ticket creation (so he does not get flooded by emails regarding tickets). - Roland points out that we need to choose a cuttof date so that a final export of the tickets is performed and no new tickets are written to the bitbucket. repo. - The conversion will have to be done over a period of 3 days, otherwise the GH API will flag us as evil and prevent it from happening. - ETK maintainers hace access to the scripts Zach is using for the migration. - Bayhan asks if the migration includes the attachments in the issues. Zach says yes. Attachments will be migrated. For now the tickets are a repository and the attachment files are *in* the repo. - Zach made it public, so that interested people can help him make the transition and test the new system. - Roland says that it would be really nice for the ticket numbers to match between GH and Bitbucket. Zach says this may be hard: What if people deleted their tickets in bitbucket and the numbering is not as sequential as one expects. Roland says he may need to create dummy tickets so the numbers line up. - Roland found that its also very hard to search for issues using the GH search bar and the original BB issue tag. # Unanswered question on mailing list - "Question about how to generate initial data with arbitrary EOS": In the body of the question, the asker specifies he is talking about BNS ID. Zach has pointed out that Sam Tootle should be the person to be talked to regarding Fuka. Also Konrad. Roland asked Beyhan to reply, he says he does not what to say. There are many ways to generate BNS ID. Roland will start the reply and others can chime in. - "Issues configuring Simfactory for PBS cluster (Hubble) - Jobs failing to submit correctly": Beyhan directed him to fix his files. - "Question about IllinoisGRMHD C2P failure and kuibit Psi4 extraction": Roland asked zach to chime in and ask for more time in the reply as he is busy and travelling. # Tickets - Looked at the most recent tickets, but had no time to discuss them in detail.

1 0

Updating the ET with GetComponents
by Jordan Nicoules 23 Jul '26

23 Jul '26

Dear all, I wanted to update the Toolkit to the new release. When I use GetComponents with the updated thornlist, it properly checks out new components. However, on existing components, it doesn't seem to add the new ET_2026_05 branch, I only still have the previous one (in that case, ET_2023_05 on my machine). This is the case even when answering yes to the update prompt. However, the script does complain, for instance, if the URL of the repo has changed. Is that normal behavior? Looking at the options for GetComponents, it's not clear to me that another option would do the job. I understand that it could mess up local modifications of repos (added branches, remotes,...), but I guess I was expecting that at least it could add the new branch, or complain for repos that may have conflicts (like for the URL). I know I could just create a new Cactus directory, or move/remove the repos for which I want an automatic update (that's what I did for the URL case), and manually update the other ones, but that sounds like it defeats the purpose of GetComponents, doesn't it? Or am I misunderstanding the role and scope of that script? Best, Jordan

2 1

Einstein Toolkit Meeting Reminder
by users＠einsteintoolkit.org 22 Jul '26

22 Jul '26

1 0

Question about how to generate initial data with an arbitrary EOS
by MAKINO Taiki 21 Jul '26

21 Jul '26

Dear Einstein Toolkit Users, I am a new Einstein Toolkit user, and I would like to ask how to generate initial data with an arbitrary neutron star EOS. My purpose is to perform binary neutron star simulations with a tabulated or an analytical EOS(for example, a BSk EOS or another tabulated EOS that I prepare myself). Could you tell me the recommended way to generate such initial data for use with Einstein Toolkit? If there is a related document or examples, I would appreciate it if you could tell me. Thank you very much for your time and help. Best regards, Taiki Makino Graduate School of Science, Nagoya University makino.taiki.g6(a)s.mail.nagoya-u.ac.jp<mailto:makino.taiki.g6@s.mail.nagoya-u.ac.jp> 取得 Outlook for Mac<https://aka.ms/GetOutlookForMac>

1 0

Issues configuring SimFactory for PBS cluster (Hubble) - Jobs failing to submit correctly
by Beyhan Karakaş 20 Jul '26

20 Jul '26

Hello Taiki, You are modifying a generic machine file and using it with pbs-normal submission file. Submission and running the simulation are done by different files and currently you use generic config and runscript (generic.cfg, generic.run). I would suggest you to use consistent simfactory files, lets say modify the files for crux (crux.ini, .cfg, .sub, .run) to work with your cluster. After having consistent files add "--machine hubble" option when you submit the job (./simfactory/bin/sim submit ... --machine hubble). Also have a look at the all supported clusters that use PBS on simfactory/mdb/submitscripts. Best regards, Beyhan. I am trying to configure SimFactory to submit jobs using PBS on a machine > named "Hubble". > > Initially, the automatically generated hubble.ini was configured for a > local login node submission. I attempted to modify the submission commands > to support our PBS environment, but the jobs are not running properly. > > Here are the modifications I made to the .ini configuration: > > * submit: Changed from: exec nohup @SCRIPTFILE@ < /dev/null > @RUNDIR@ > /@SIMULATION_NAME@.out 2> @RUNDIR@/@SIMULATION_NAME@.err & echo $! To: > qsub @SCRIPTFILE@ > > * stop: Changed from: pkill -g $(ps -o pgid= -p @JOB_ID@) To: qdel > @JOB_ID@ > > * getstatus: Changed from: ps @JOB_ID@ To: qstat -f @JOB_ID@ > > Symptoms: > > 1. When running simfactory/bin/sim submit, no .out or .err files are > generated at all. > > 2. When running simfactory/bin/sim create-submit, the .out and .err > files are generated, but the .out file only contains the PBS job ID, and > the .err file is completely empty. > > 3. No simulation output files or directories are created inside the > designated simulation directory. > > As you can see from the symptoms, create-submit successfully captures the > job ID, but the simulation does not run and no output is generated. > > Could anyone guide me on the correct way to write the submit option or the > template script for PBS in SimFactory? Any working examples of a PBS > configuration (.ini and .sub files) would be greatly appreciated. > > For your reference, I have attached my .ini file and .sub file that I am > currently using. > > >

2 1

← Newer
1
2
3
4
5
6
7
8
...
399
Older →

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users