Users September 2020

users@lists.einsteintoolkit.org

18 participants
19 discussions

WG: Top for next weekly meeting
by Schaffarczyk, Alois 14 Sep '20

14 Sep '20

* 3rd attempt -- Dear all, My mail from Sept. 4 attached an improved plot (hopefully) obeying all recommendations from Sept-3: 1. Used color palette "virbis" 2. NR template mass scaling was adapted to value as stated in BBH gallery URL (0.95) My mail from Sept. 3 This is our (Atul and me) wish: Add the attached figure close to the trajectory. Justification: 1. Shows almost exact agreement to former ETK version, 2. Shows almost exact agreement to NR template, 3. Leads directly to the experiment it was designed for. We ask for your kind consideration Best wishes Peter S, Kiel Germany

1 0

ReadInterpolate thorn final review
by Roland Haas 10 Sep '20

10 Sep '20

Hello all, the ReadInterpolate thorn which is a FileReader like thorn that uses InterolateLocalUniform to interpolate the data read in onto the new grid is being proposed for inclusion in the next Einstein Toolkit release ET_2020_11. It has been positively reviewed in: https://bitbucket.org/einsteintoolkit/tickets/issues/2416 and all are invited to have a look at it and all in to the next weekly Einstein Toolkit call on Thu Sept 17th 2020 to cast their vote about it being accepted into the toolkit. The thorn has been part of the master thorn list and been included in the automated tests since Fri Jul 17 2020. The ticket provides a link to the source of the thorn in its description. 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

Error in a function
by Erik Rodrígo Jiménez Vázquez 09 Sep '20

09 Sep '20

Hello everybody, I'm writing a function to set initial conditions, but I need to use an interative method and I need to operate over some arrays (grid functions) that I previously declared in the interface.ccl file. This is an example of one of the functions that I'm currently writing. #include "cctk.h" #include "cctk_Arguments.h" #include "cctk_Parameters.h" #include "ID.h" void Initial_Setup(CCTK_ARGUMENTS) { DECLARE_CCTK_ARGUMENTS; DECLARE_CCTK_PARAMETERS; int index; int i,j,k; int ierr; for (k=0; k<cctk_lsh[2]; k++) { for (j=0; j<cctk_lsh[1]; j++) { for (i=0; i<cctk_lsh[0]; i++) { index = CCTK_GFINDEX3D(cctkGH,i,j,k); phi[index] = 1.0; pi [index] = 0.0; } } } } where pi and phi are the grid functions previously defined in the interface.ccl and the header "ID.h" is where I defined the function Initial_Setup When I call the function in the main function and compile it I get error: type name is not allowed Initial_Setup(CCTK_ARGUMENTS); error: expression must have arithmetic type Initial_Setup(CCTK_ARGUMENTS); these errors are the same for the other functions. All these functions are not scheduled but the main function is. What is the problem? Any help will be apreciated. Erik

3 2

Midwest Relativity Meeting 2020, October 22-24
by Atul Kedia 08 Sep '20

08 Sep '20

Dear Relativists, The 30th annual Midwest Relativity Meeting will be hosted by the Center for Astrophysics and the Department of Physics of the University of Notre Dame from *October 22-24*, Thursday to Saturday. The Midwest Relativity Meetings aim to bring together researchers from across the Midwest and* beyond* to discuss general relativity and a broad range of topics in gravitational physics, including classical and quantum gravity, numerical relativity, relativistic astrophysics, cosmology, gravitational waves, and experimental gravity. As was in the past, talks will consist of short contributed papers in an informal atmosphere. Each participant may present a talk of ~15 minutes. Professors, post-docs, graduate students, and undergraduates working on relevant topics are encouraged to present their work. The deadline for abstract submission is *9th October*. The Blue Apple Prize will be awarded to the best graduate student speaker. Venue: Due to COVID-19 travel restrictions we will not be able to host the meeting at our beautiful campus. Rather, we will be conducting the meeting virtually. Recent experience has indicated that it is difficult to participate in full days of talks in a virtual format. Hence, we will extend the meeting over an additional day, and limit sessions to *no more than 4 hours per day* in the afternoons. Registration: To register and propose a talk, please visit the conference web site at: https://sites.google.com/nd.edu/mwrm-2020 Registration Fee: There is no registration fee. We look forward to seeing you at the 2020 Midwest Relativity Meeting. Sincerely, Atul Kedia PhD student, University of Notre Dame On behalf of the Local Organizing Committee: Grant Mathews, Lan Nguyen, Arielle Phillips, In-Saeng Suh, Atul Kedia, Miguel Correa, Luca Boccioli, Xilu Wang. Scientific Organizing Committee: Grant Mathews – U. Notre Dame David Garfinkle – Oakland Univ John Friedman – U. Wisc. Milwaukee Brett Bolen – Grand Valley State Univ. (last year’s host) Nicholas Yunes – UIUC (next year’s host)

1 0

Agenda for Thursday's Meeting
by users＠einsteintoolkit.org 07 Sep '20

07 Sep '20

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

1 0

PostCactus installation.
by Bilal Hameed 06 Sep '20

06 Sep '20

Dear users, I am installing "PostCactus" as $ pip install PostCactus. My python version is, *Python 2.7.17* Still, I am getting following error, Defaulting to user installation because normal site-packages is not writeable ERROR: Could not find a version that satisfies the requirement PostCactus (from versions: none) ERROR: No matching distribution found for PostCactus

2 1

Compilation error in PointerTo.F90
by Chris Stevens 05 Sep '20

05 Sep '20

Hi everyone, I am trying to get the Einstein Toolkit compiling on a new cluster, and most of the code compiles but I get the following error: COMPILING Cactus/util/PointerTo.F90 /maybehome/cstevens/Cactus_For_BBH_Production_Runs/Cactus/src/util/PointerTo.F90:23.27: address = transfer(c_loc(var), address) 1 Error: Assumed-rank argument requires an explicit interface at (1) make[3]: *** [PointerTo.F90.o] Error 1C make[2]: *** [util/make.checked] Error 2 make[1]: *** [/maybehome/cstevens/Cactus_For_BBH_Production_Runs/Cactus/configs/sim/lib/libthorn_Cactus.a] Error 2 I am using gfortran 4.8.5, perhaps this can be resolved by simply updating gcc and gfortran? The flags for Fortran are F90FLAGS = -g3 -fcray-pointer -ffixed-line-length-none Cheers, Chris -- Dr Chris Stevens Postdoctoral Fellow Rm 310A Department of Mathematics and Statistics University of Otago Ph: +64 3 479 4567 http://www.chrisdoesmaths.com http://gravity.otago.ac.nz

2 2

Meeting Minutes 2020-09-03
by Gabella, William E 03 Sep '20

03 Sep '20

Present: Bill G, Yosef Z, Roland H, Ken S, Peter D, Zach E, Steve B, Peter S, Federico C, Atul Chair: Yosef Minutes: Bill [short agenda] * Peter S, request to add another plot for the Binary Black Hole example. Yosef, what about the ringdown ramping up at the end? Peter, just used the POWER routines. Choice of mass determines the timescale. Roland, M_Adm early is 1.0, or after GWs pass is 1.0. Either choice is defensible. LIGO and ET make the opposite choice. Style: Be Sensitive to blue-green color blindness; can google for color pallettes. Recent gnuplot (4 and beyond?) has good color palettes for that. * Steve, we have a replacement for cactuscode.org. Current one is on old system and full of stuff that no one wants to maintain. No one liked making it a sub-page off of ET. Ian Hinder built a version on GitHub on Jekyll and has some things missing. Need smaller, easier to maintain website. Yosef, if it is good enough and make it live and let the complainers help fix it. Steve, we should tweak a couple of things: there are a bunch of tickets that talk about what are wrong with it. Some issues to fix to make it live. https://test.cactuscode.org https://github.com/EinsteinToolkit/cactuscode.org/issues?q=is:open+is:issue… * Unanswered Questions mailing list o No one waiting for help. * Open tickets sorted by update time o Roland, #1717 from 2014, about libraries and automatic build which does not pick them up. Suggestion is to build dynamically, but then harder for systems that need static library building (Cray). Proper solution is to have the actual build stage of the libs return build variables. o #2448 remove the support_old_CalcTmunu_mechanism interface to Tmunu, probably the only thing that is visible to people. * Tickets ready for review Any comments? o Roland, a lot have existed for a while. Tmunu mentioned above. #2449 NewRad extrapolate the Gammas (Christoffel) from data interior to the boundary, there is code that grid has min size in all three Cartesian dimensions but fails with Cartoon2D because it will not have that many points (one interior)...because symmetry conditions. This should not do anything for a real run...you need five points. Only effects the extrapolation. o Roland, #2458 Simfactory Done for the correct date. These have a Pull Request and need someone to review the Pull Request. oo Current "policy": More things up for review than people willing to review them, typically if they are there for a while, 2 weeks put in "any suggestions" marker and after another two weeks tey are Pushed in the code. Pure enhancements are not treated this way. They need real second pair of eyes checking them out. o Yosef, should we assign reviewers? Roland, sure, I often give them to Steve to look at. A lot fo these are one line change. Bill will review #2458, Simfactory Done with time. Steve, #2389 make error messages better for checkpoint files, this looks good to him. Steve logged in to the Pull Request and approved the changes. * Working group reports o Zach's IllinoisGRMHD WG update: IllinoisGRMHD is getting a more robust conservative-to-primitives solver, more along the lines of what HARM does. Basically IllinoisGRMHD evolves the entropy as an additional conservative variable, and uses this new variable when the usual conservative-to-primitive solver fails. This adds robustness to modeling highly magnetized, relativistic fluid flows. We'd like to make this update public in a month or so... everything is implemented and we're debugging now. o No other groups reporting. * Other business o Ken Sibel, post-bac researcher with Zach, and will be applying to grad school for CS. Did all the work on the early version on NRPy LaTeX interpreter. Next week may have latest bits in place. REady for testing/experimentation in notebooks. Insert LaTeX equations and will output symbolic expressions for Mathematica and for Pythone etc. Next project is GPU support in NRPy. o Atul, announcement of the virtual conference "Midwest Relativity Meeting 2020"---is the 30th year, Oct 22-24, 2020. Hosted by Notre Dame. Topics from grav physics, num rel, rel astrophysics, cosmology. Abstract submission is open. Can I send an email to the user list. * People to run the call... o Have next weeks volunteers. o Sept 17th: chair Peter D, minutes Steve -- ===================================== William Gabella Research Assistant Professor Department of Physics and Astronomy Vanderbilt University Nashville, TN USA b.gabella(a)vanderbilt.edu (o) 615-343-2713

1 0

.h5 file issue.
by Bilal Hameed 01 Sep '20

01 Sep '20

Dear Users, I am trying to open a .h5 file in jupyter-notebook. Please anyone guide me about reading of HDF files. *_Command;_* import h5py filename = "ct_multilevel-err.h5" with h5py.File(filename, "r") as f: # List all groups print("Keys: %s" % f.keys()) a_group_key = list(f.keys())[0] # Get the data data = list(f[a_group_key]) *_Error;_* IOError Traceback (most recent call last) <ipython-input-12-769b4593257a> in<module>() 1 filename= "ct_multilevel-err.h5" ----> 2with h5py.File(filename, "r") as f: 3 # List all groups 4 print("Keys: %s" % f.keys()) 5 a_group_key= list(f.keys())[0] /usr/local/lib/python2.7/dist-packages/h5py/_hl/files.pyc in__init__(self, name, mode, driver, libver, userblock_size, swmr, rdcc_nslots, rdcc_nbytes, rdcc_w0, track_order, **kwds) 406 fid = make_fid(name, mode, userblock_size, 407 fapl, fcpl=make_fcpl(track_order=track_order), --> 408swmr=swmr) 409 410 if isinstance(libver, tuple): /usr/local/lib/python2.7/dist-packages/h5py/_hl/files.pyc inmake_fid(name, mode, userblock_size, fapl, fcpl, swmr) 171 if swmrand swmr_support: 172 flags|= h5f.ACC_SWMR_READ --> 173fid= h5f.open(name, flags, fapl=fapl) 174 elif mode== 'r+': 175 fid= h5f.open(name, h5f.ACC_RDWR, fapl=fapl) h5py/_objects.pyx inh5py._objects.with_phil.wrapper() h5py/_objects.pyx inh5py._objects.with_phil.wrapper() h5py/h5f.pyx inh5py.h5f.open() *IOError**: Unable to open file (unable to open file: name = 'ct_multilevel-err.h5', errno = 2, error message = 'No such file or directory', flags = 0, o_flags = 0)*

3 2

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users September 2020