Users May 2021

users@lists.einsteintoolkit.org

16 participants
26 discussions

Agenda for Thursday's Meeting
by users＠einsteintoolkit.org 17 May '21

17 May '21

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

1 0

Re: [Users] Query regarding visualisation software/packages
by Gabriele Bozzola 17 May '21

17 May '21

Dear Annamalai, I don't know much about yt, so the help I can offer is limited and I can only point out what you need to be able to do with yt to make a video like the one I made. I personally think that yt is not the correct tool to use for this kind of plots because you need to be able to draw several kinds of structures (curves, arrows, meshes, volume rendering), and I am not sure what is the level of support for all these different kinds of plots in yt. I used mayavi for the example. 1. Horizons are meshes (typically, 6 partially overlapping meshes). So, you need to figure out how to draw meshes with yt. You can learn about how these are represented by kuibit in the tutorial on horizons. (https://sbozzolo.github.io/kuibit/tutorials/cactus_horizons.html). I report here one of the figures to give an idea of how horizons are represented . [image: image.png] You can obtain the 3D position of all the vertices with the function `shape_at_iteration` or `shape_at_time`. 2. Spins are vectors. You have to find out how to draw arrows in yt. You can use the function `ah_origin_at_iteration` to find the center of the horizon at any given time, and then you can find the value of the vector you are interested in simply getting the quantity with brackets, e.g. h["momentum_x"] would give you a timeseries with the value of the linear momentum along the x direction for horizon h. h["momentum_x"](time) is the value of that component at the given time, which you can use to draw the arrow. 3. Orbits are 3D curves. h.ah.centroid_x is a timeseries with the x coordinate of the horizon h. You can use this to draw a 3D plot with the orbits. Read tutorial on timeseries for more information (https://sbozzolo.github.io/kuibit/tutorials/timeseries.html) 4. The curvature is the only quantity for which I think it makes sense to plot with yt. It is a 3D grid function. Normally, this has all the information about the refinement levels, but you can resample everything to a uniform grid. This can be passed to yt as "generic data", as shown here https://yt-project.org/doc/examining/generic_array_data.html. For information about grid data, have a look at the relevant tutorial https://sbozzolo.github.io/kuibit/tutorials/grid_data.html. When you use mayavi, this can be a 3D contour plot, but it still has to be resampled onto an uniform grid. 5. Gravitational waves are 2D curves that live on a different space compared to the main plot. I used matplotlib to draw this (plt.plot(strain)) and then joined the images with Pillow. The tutorial on gravitational waves will tell you more about this: https://sbozzolo.github.io/kuibit/tutorials/gravitational_waves.html 6. If you want to have something nice and user-friendly, you can take inspiration from the examples in kuibit. These examples are command-line utilities that work out-of-the-box for all the simulations. They have a lot of options and other nice features (e.g. tab-completion). Some documentation about that is here https://sbozzolo.github.io/kuibit/argparse_helper.html, but the examples themselves are great starting points. My personal recommendation for how you should move forward is to familiarize yourself with kuibit first. In this email I sketched how you can achieve what you want, but I neglected a ton of details. I suspect that it is only after you learned the basics of kuibit that this email will make sense. Also, you should ask yourself if you really want to use yt for this specific use case. If you don't want to, everything I described in this email is already implemented (with mayavi) and you can simply customize the example that I pointed out in the previous email for your needs. To get started with kuibit, you can read the tutorials, https://sbozzolo.github.io/kuibit/#tutorials and this page https://sbozzolo.github.io/kuibit/simdir.html. Then, I'd suggest trying to re-implement one of the examples on your own, for example plot_ah_separation.py ( https://sbozzolo.github.io/kuibit/examples/bins/plot_ah_separation.html). With this, you will gain enough experience to move to the next step and make the movie you want to. Every single function in kuibit is documented. In addition to the examples and tutorials, we have two additional types of documentation (for a total of four different types): - Explanations of the various modules ( https://sbozzolo.github.io/kuibit/#usage), where you can see how the modules should be used and a broad overview on how they are structured and what features they provide. - Reference and technical details of the APIs ( https://sbozzolo.github.io/kuibit/#reference-material-classes-functions), where you can see the details of the function signatures, the types expected, et cetera. You will likely need the latter two forms of documentation when you will start writing code with kuibit. Best, Gabriele On Mon, May 17, 2021 at 11:05 AM Annnamalai P S <annamalaips.97(a)gmail.com> wrote: > Dear Gabriele, > > Thank you for your reply and for taking the time to point out the > resources you felt would be relevant to my current research. > > I'm currently not very well versed with yt package but am very much > interested in the idea that you have suggested. But at the same time I'm > new to kuibit as well. If you can share any resources/documentation that > helps me to prepare data to be used for yt package that would be great. I > hope to create animated volume rendering using yt package like the ones you > have created. I also want to set it up as a user friendly and out of the > box package that can yield plots for visualization. > > On Mon, May 17, 2021 at 8:52 PM Gabriele Bozzola < > bozzola.gabriele(a)gmail.com> wrote: > >> Dear Annamalai, >> >> I am the author of kuibit, and I would like to highlight that the package >> is capable of doing what you are looking for, >> including 3D rendering. The package also comes with a lot of examples >> that you can immediately use for your >> goals. There are already some examples for 2D/3D plots of binary black >> holes, and you can see an example of >> rendering here: >> https://github.com/Sbozzolo/kuibit/tree/experimental/examples >> where I rendered a 3D movie with the apparent horizons, magnitude of the >> spins, orbits, and gravitational wave >> strain. You can produce the same movie for your simulation writing 0 >> lines of code [1]. >> >> As for yt, kuibit can prepare the data so that it is trivial to use it >> with yt (see GitHub issue >> https://github.com/Sbozzolo/kuibit/issues/18). I've never used yt and I >> don't even know where to start, but if you >> are already familiar with it, you'll be able to use it with kuibit in no >> time. [2] >> >> Finally, let me point to the documentation: >> https://sbozzolo.github.io/kuibit/. kuibit is thoroughly documented and >> there are lots of examples and tutorials that will help you get started. >> In addition to that, we have a kuibit user >> group chat on Telegram (https://t.me/kuibit) where you can ask questions >> if you find that something is not clear. >> >> Let me know if you have any questions, >> Gabriele >> >> >> [1] The code is in the experimental branch, so I would actually not >> recommend using it at the moment because it >> will break in the future. However, you can still see how it is done there >> and use the stable version of kuibit to >> achieve a similar result. At this point in time, I would recommend NOT >> using the experimental branch. >> >> [2] In that case, we would love you to contribute back to kuibit :) >> >> On Mon, May 17, 2021 at 8:05 AM Roland Haas <rhaas(a)illinois.edu> wrote: >> >>> Hello Annamalai, >>> >>> please take a look at our wiki: >>> >>> https://docs.einsteintoolkit.org/et-docs/Analysis_and_post-processing >>> >>> Linked in the "Analyzing Data Output by the Einstein Toolkit" >>> bullet of "Documentation" section of the wiki at >>> >>> https://docs.einsteintoolkit.org/et-docs/Main_Page >>> >>> Please note that yt support is not currently functional since the >>> repository linked to no longer exists. >>> >>> The most functionally complete packages are likely: >>> >>> * SimulationTools >>> * PostCactus >>> * Kuibit >>> >>> For 3d visualization I think you are pretty much limited to VisIt >>> unless you convert the HDF5 files into a format understood by other >>> packages. >>> >>> Yours, >>> Roland >>> >>> > Hi Einstein Toolkit users, >>> > >>> > I'm a relatively new Einstein Toolkit user learning to simulate and >>> > visualise binary black hole system mergers. I'm trying to write a code >>> to >>> > visualise animated trajectories, horizons, gravitational waveforms and >>> > curvature of binary system merger using Python matplotlib and >>> yt-project >>> > packages. >>> > >>> > Before I proceed, I want to know what are the list of available >>> > open-source/commercial visualisation software/packages that can produce >>> > animated plots(2D and 3D) of binary system merger with data from >>> Einstein >>> > toolkit output, and get an idea of whether something similar already >>> exists. >>> > >>> > >>> > -- >>> > Best regards, >>> > Annamalai P S >>> >>> >>> -- >>> My email is as private as my paper mail. I therefore support encrypting >>> and signing email messages. Get my PGP key from http://pgp.mit.edu . >>> _______________________________________________ >>> Users mailing list >>> Users(a)einsteintoolkit.org >>> http://lists.einsteintoolkit.org/mailman/listinfo/users >>> >> > > -- > Annamalai P S >

2 1

Query regarding visualisation software/packages
by Annnamalai P S 17 May '21

17 May '21

Hi Einstein Toolkit users, I'm a relatively new Einstein Toolkit user learning to simulate and visualise binary black hole system mergers. I'm trying to write a code to visualise animated trajectories, horizons, gravitational waveforms and curvature of binary system merger using Python matplotlib and yt-project packages. Before I proceed, I want to know what are the list of available open-source/commercial visualisation software/packages that can produce animated plots(2D and 3D) of binary system merger with data from Einstein toolkit output, and get an idea of whether something similar already exists. -- Best regards, Annamalai P S

4 3

HelpEinsteinToolkit: "required key: user"
by jumipas＠alumni.uv.es 17 May '21

17 May '21

Dear users of Einstein Toolkit, I'm a pregraduated student in University of Valencia (Spain) and I'm using Einstein Toolkit for the first time (I used the stand-alone version of RNS inside ET but I had not big problems with it) as part of a schollarship I got this year. When I tried to run a simulation, ET returned an error which said that "consts.h" (located in ET/Cactus/arrangements/EinsteinInitialData/Hydro_RNSID/src/include) was missing. However, this file actually exists and it is located in that same directory. The error was something like this: rnsid_util.o fatal error #include "consts.h" The file consts.h does not exist or cannot be read. It was not exactly like this message, but it was quite similar. After many tries without solving this problem, I tried reinstalling the Einstein Toolkit (now I have 2) and it did not appear again, so it might be an installation error or something similar. However, now ET returns another error (actually, the machine in which I'm working is the one which returns the error, but it is from something missing in ET, not the machine). The error it returns is: Error: machine [name] is missing a required key: user Aborting Simfactory. And I have tried almost everything I know to solve it. I also spoke to some colleagues here that also use ET but all the files we compared were identical (except for the initial path to Einstein Toolkit and the username). We thought it could have something to do precisely with the username, but we changed it several times and we could not fix the error. As I still have both versions of ET (one with each error), any help with each of the errors will be much appreciated. Thank you in advance. Sincerelly, -J.

2 3

meeting minutes for 2021-05-13
by Roland Haas 13 May '21

13 May '21

Present: Roland, Peter D., Maria, Atul, Ken, Leo, Miguel G., Bill G., Zach Gallery examples: * BBH: no update, no Steve * TOV: Peter ran the example, will update the gallery page, failure with presync_mode = "mixed-errors" persists, will fix after release * Multipatch: was run successfully, Bill working on visualization, ran fine with presync_enabled, will check that results are unchanged after feature freeze * BNS: Atul ran into issues that svn.cct.lsu.edu for is no longer accessible (for NSTracker), happened in the last couple of weeks, tested right now with offline copy of NSTracker (80% run done). Need to find alternative location for NSTracker. Maria mentioned ability to run without NSTracker by using larger box. 3 options are possible: NSTracker, or WVUThorns/VolumeIntegrals (cannot handle symmetry), box big enough to cover full orbit of star. Could also advect a tracer at center of star, with ETTracer, and naive implementations have the same problem with symmetries. Would want to include instructions on how to relax symmetry in gallery parfile. Atul will try and see if he can modify the parfile to not use symmetry, may add code to VolumeIntegrals_GRMHD to handle z-symmetry. Example parfiles are in repos/wvuthorns_diagnostics/NSNS_parameter_files * Poisson equation: worked with presync, no problems and results look the same as last time * POWER: Roland adding some more features to handle cutoff frequency and ADM mass, still needs documentation * tickets flagged as major: ** Lean, ignored for now, https://bitbucket.org/einsteintoolkit/tickets/issues/2487 ** ExternalLibraries, has workaround, ticket has a suggested solution, https://bitbucket.org/einsteintoolkit/tickets/issues/2528 ** TmunBase in McLachlan and presync, will look into after release, https://bitbucket.org/einsteintoolkit/tickets/issues/2526 ** WVUThorns_Diagnostics documentation, Leo will look into this, https://bitbucket.org/einsteintoolkit/tickets/issues/2469 ** different results on different number of cores, Atul may have seen a similar effect on some of his own simulations, Maria might have seen something similar that was fixed by changing some grid structures, will check https://bitbucket.org/einsteintoolkit/tickets/issues/2516 ** only real blocker right now is POWER's incompleteness, where we want to emphasize at first release that it has undergone rigorous review and is useful, want to make sure that all gallery example ticket have been updated and still work Yours, Roland -- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://pgp.mit.edu .

1 0

Einstein Toolkit Meeting Reminder
by users＠einsteintoolkit.org 12 May '21

12 May '21

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

SimulationTools - loading the grid functions
by Konrad Topolski 12 May '21

12 May '21

Hi, I was wondering if anyone could help me with loading the data from the simulation via the SimulationTools. I have confirmed that there is data available, but cannot load it to a variable via ReadGridFunction. Below I'm copying the notebook's content up to the point where I cannot load the data. *SimulationTools`$SimulationPath = {"C:\\Users\\Seven\\Desktop\\BBH_run_01"};* *FindGridFunctions["BBH_run_01"]* "admbase-curv", "admbase-lapse", "admbase-metric", \ "carpetreduce-weight", "grid-coordinates", \ "ml_admconstraints-ml_ham", "ml_admconstraints-ml_mom", \ "weylscal4-psi4i_group", "weylscal4-psi4r_group" *ReadIterations["BBH_run_01", "admbase-lapse", "xy"]* {0, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2304, 2560, 2816, \ 3072, 3328, 3584, 3840, 4096, 4352, 4608, 4864, 5120, 5376, 5632, \ 5888, 6144, 6400, 6656, 6912, 7168, 7424, 7680, 7936, 8192, 8448, \ 8704, 8960, 9216, 9472, 9728, 9984, 10240, 10496, 10752, 11008, \ 11264, 11520, 11776, 12032, 12288, 12544, 12800, 13056, 13312, 13568, \ 13824, 14080, 14336, 14592, 14848, 15104, 15360, 15616, 15872, 16128, \ 16384, 16640} *ReadRefinementLevels["BBH_run_01", "admbase-lapse", "xy"]* {0, 1, 2, 3, 4, 5, 6} *ReadTimeLevels["BBH_run_01", "admbase-lapse", "xy"]* {0} *alpha = ReadGridFunction["BBH_run_01", "admbase-lapse", "xy", Iteration -> 4096, RefinementLevel -> 0]* Cannot find a variable named admbase-lapse in files {C:\Users\Seven\Desktop\BBH_run_01\admbase-lapse.xy.h5}; variables found are {ADMBASE::alp} in ReadGridFunction/ CallProvidedFunction/ SimulationTools`CarpetIOHDF5`GridFunctions`ReadData/ SimulationTools`CarpetIOHDF5`Private`ReadCarpetIOHDF5Components *Trying to load ADMBASE::alp yields: * Error: Iteration 4096, RefinementLevel 0, Map None of variable ADMBASE::alp with dimensions xy not found in run BBH_run_01 The relevant file I want to load is in /BBH_run_01/admbase-lapse.xy.h5 What's the solution here? What am I doing wrong? Best regards Konrad Topolski

2 3

Installation Problem
by Jaime Redondo Yuste 12 May '21

12 May '21

Hello there! I am new to all of this, so let me apologize if this questions has been already asked or if its solution is obvious: I have done some search and I have seen this question has been asked before (https://bitbucket.org/einsteintoolkit/tickets/issues/2069/connection-failur… <https://bitbucket.org/einsteintoolkit/tickets/issues/2069/connection-failur…>) but I cannot understand what should I do to fix it. I have never worked with SVN so I understand, basically, nothing. The problem is, when I try to download the ET, most of the packages install correctly, but some of them deliver this error: Could not checkout module ExternalLibraries/hwloc svn: E170013: Unable to connect to a repository at URL 'https://github.com/EinsteinToolkit/ExternalLibraries-hwloc.git/branches/ET_…' svn: E120171: Error running context: An error occurred during SSL communication I have tried installing again SVN, this is my current version: svn, version 1.11.1 (r1850623) Also I am working on a MacBook (10.15.6), if that is of any help. I have no idea what to do to fix this. I would appreciate any help or guidance that you might provide. Thank you all in advance, I am so excited to join this community! Best, Jaime

2 1

Agenda for Thursday's Meeting
by users＠einsteintoolkit.org 10 May '21

10 May '21

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

1 0

Meeting Minutes 2021-05-06
by Gabella, William E 06 May '21

06 May '21

Present: Bill G, Steve B, Peter D, Roland H, Ken S, Atul K, Zach E, Miguel G, Leonardo W, Maria H Chair: Steve B, Minutes: Bill G * Workshop update Online hosted by UIUC on July 26th to 30th, 2021. https://einsteintoolkit.github.io/et2021uiuc/ Roland, have to announce for registration, have everything ready to run it. First meeting of Scientific Organizing Committee to hash out the program, contacting potential speakers. Email(s) forthcoming. * Release Planning [https://docs.einsteintoolkit.org/et-docs/Release_Details Release Details] ** Gallery examples: *** [https://bitbucket.org/einsteintoolkit/tickets/issues/2172/test-binary-black… BBH] [SB] Steve B, Presync, with error mode turned on it does not work. Cannot synchronize past time levels. *** [https://bitbucket.org/einsteintoolkit/tickets/issues/2175/test-single-stabl… TOV] [PD] Peter D, TOV was fine. Turning on Presync checking and it failed, looks like Machlachlan has incorrect read declarations. Checked git Log for Kranc and found where the read statements were put in. Roland commented them out by hand---Roland, we try not to do that. Now they are there and uncommented. Added a ticket [https://bitbucket.org/einsteintoolkit/tickets/issues/2526/mclachlan-has-inc… #2526], assigned to Roland. After commenting them out by hand it does work with Presync. *** [https://bitbucket.org/einsteintoolkit/tickets/issues/2174/test-multi-patch-… Multipatch] [BG] Bill G, Multipatch seems to work with both original par file and with Presync. Need to get the Visit session plotting. *** [https://bitbucket.org/einsteintoolkit/tickets/issues/2176/test-binary-neutr… BNS] [AK] Atul K, BNS merger and ran it, listed as 8 cores 24 hours, and used 2 cores took 3 days. Seems fine. Visit looks okay. Presync did not work, will put it on BitBucket issue. *** [https://bitbucket.org/einsteintoolkit/tickets/issues/2173/test-poisson-equa… Poisson equation] [MG] Miguel G, Poisson looks like it is running with and without PreSync, and same results as previous release. Steve B, We should keep track more formally of the runtime and machines for gallery examples. Zach E, in the running of our testsuite we do keep track of times, or should, and that should catch changes in the code. Maybe YouTube videos for tutorials on the Gallery examples...Steve, maybe from recordings at workshops or schools. ** POWER [https://bitbucket.org/einsteintoolkit/tickets/issues/2413/include-power-cod… status] 1234567890123456789012345678901234567890123456789012345678901234567890123 Zach E, POWER is our killer feature. Maria tried to use POWER and there were some hiccups. User interface could be better. Roland, there needs to be documentation but some difficulties are design decisions. Gabriele and Maria were using it as a command line tool, and it has limited flexibility. Use it as a Python module, and designed for this. IPython and write scripts. It is a single file. CLI assumes you have data laid out as in Einstein Toolkit. Roland, will not try for any auto-detect of the data layout. No tests for it yet. Zach, Prioritize documentation and later features. Roland, this method will likely be incorporated into Kuibit or other tools being developed. ** Atul K, Ran test suite on their cluster about 90 say "unrunnable"---not failed. Roland, some might not be included or compiled thorns, maybe GPU thorns, etc. Do not expect all the tests to run. ** Zach, have to prepare the release announcement. How to list all the maintainers? On the DOI have all the people that contributed to Toolkit and Cactus. New code listed, for commits that are done. Release Details page should list that. New features, and fixed bugs, list as major change. ** Name for release, we are using a male scientist this time. Look at the list. Think about it over this next week. List is in Release Details https://docs.einsteintoolkit.org/et-docs/Release_Details#Tools.2FLinks . ** [https://bitbucket.org/einsteintoolkit/tickets/issues?kind=bug&priority=majo… bugs of severity major or worse] ** [https://bitbucket.org/einsteintoolkit/tickets/issues?milestone=ET_2021_05&s… open tickets] flagged for the next release Steve B, Unanswered Questions on the mailing list seem handled. Tickets for review. Re-run galleries after the Feature Freeze, 14 May 2021. Zach, Update Docker images? Steve usually uses those for testing various Linux flavors. Tutorials may be running in Docker but they actually download the Toolkit. 13 May, Chair Maria B, Minutes Roland H 20 May, Chair Peter D, Minutes Maria B [https://www.einsteintoolkit.org/tools/unanswered.php unanswered question on mailing list] [https://bitbucket.org/einsteintoolkit/tickets/issues?status=open&status=new… open tickets sorted by update time] [https://bitbucket.org/einsteintoolkit/tickets/issues?status=new&status=open… tickets ready for review] *AOB -- ============================== William Gabella Research Assistant Professor Department of Physics and Astronomy Nashville, TN USA (o) 615-343-2713

1 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users May 2021