- Comp-phys-alps-users - lists.phys.ethz.ch

Making the scheduler switch tasks, obtaining MPI rank
by Peter Bröcker 19 Sep '13

19 Sep '13

Hello, I am using the ALPS libraries for scheduling and observable handling. I typically have more tasks than cores, especially when debugging on my personal computer. Is it possible to make the scheduler start say the first four jobs and when it's time to write a checkpoint switch to the next four instead of continuing with the first until they are finished? Also for debugging purposes, I print out some messages along the way and it would help to know which process printed them. Is the MPI rank saved somewhere or do I have to obtain it myself? Best regards, Peter

3 4

Re: [ALPS-users] Time propagation in TEBD stops prematurely
by Joseph Prentice 18 Sep '13

18 Sep '13

Hi Michael, Unfortunately, the fix you suggested, deleting one of my quenches, only worked for a little bit (although I had to change it a little, by making POW 1 instead of 0 - otherwise the quench did not occur at all). As I've increased Chi up to around 800, I've encountered the same issue as before - namely that the program just stops at around t=5 in the real time propagation when it should continue to t=15, with no error message at all. The parameters of my new Python script, incorporating the suggestion from before is below. Is there anything else I can do? I'm running on a Linux cluster. Any help would be brilliant. Thank you very much, Joseph Python script parameters: parms = [ { 'L' : 128, 'MODEL' : 'spin', 'local_S' : 0.5, 'CONSERVED_QUANTUMNUMBERS' : 'Sz', 'Sz' : 0, 'Jxy' : 1, 'Jz' : 3.0, 'ITP_CHIS' : [80, 80], 'ITP_DTS' : [0.05, 0.025], 'ITP_CONVS' : [1E-10, 1E-11], 'INITIAL_STATE' : 'ground', 'CHI_LIMIT' : 800, 'TRUNC_LIMIT' : 1E-12, 'NUM_THREADS' : 1, 'TAUS' : [15.0], 'POWS' : [1.0], 'GS' : ['Jz'], 'GIS' : [6.0], 'GFS' : [6.0], 'NUMSTEPS' : [625], 'STEPSFORSTORE' : [10], 'VERBOSE' : 'true' } ] On 22 Aug 2013, at 11:00, <comp-phys-alps-users-request(a)lists.phys.ethz.ch> <comp-phys-alps-users-request(a)lists.phys.ethz.ch> wrote: > Send Comp-phys-alps-users mailing list submissions to > comp-phys-alps-users(a)lists.phys.ethz.ch > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.phys.ethz.ch/listinfo/comp-phys-alps-users > or, via email, send a message with subject or body 'help' to > comp-phys-alps-users-request(a)lists.phys.ethz.ch > > You can reach the person managing the list at > comp-phys-alps-users-owner(a)lists.phys.ethz.ch > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Comp-phys-alps-users digest..." > > > Today's Topics: > > 1. Alps 2.1.2 r6963 hybridization expansion: No parameter > available (backes(a)th.physik.uni-frankfurt.de) > 2. Time propagation in TEBD stops prematurely (Joseph Prentice) > 3. Re: Alps 2.1.2 r6963 hybridization expansion: No parameter > available (Emanuel Gull) > 4. Re: Alps 2.1.2 r6963 hybridization expansion: No parameter > available (Hartmut Hafermann) > 5. Re: Time propagation in TEBD stops prematurely (Michael Wall) > 6. Re: Alps 2.1.2 r6963 hybridization expansion: No parameter > available (backes(a)th.physik.uni-frankfurt.de) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 21 Aug 2013 13:51:25 +0200 > From: backes(a)th.physik.uni-frankfurt.de > To: comp-phys-alps-users(a)lists.phys.ethz.ch > Subject: [ALPS-users] Alps 2.1.2 r6963 hybridization expansion: No > parameter available > Message-ID: > <70b90575915bdffa61ec46f7481c0ecb.squirrel(a)th.physik.uni-frankfurt.de> > Content-Type: text/plain;charset=utf-8 > > Dear all, > > I want to use the newest version of the ALPS Hybridization Expansion > Impurity Solver from the nightly build 2.1.2-r6963. > Compilation is successful, but when running the solver with a simple test > case I get the following Error (I shortened the paths for readability): > >> hybridization parm.h5 > > terminate called after throwing an instance of 'std::runtime_error' > what(): No parameter available > In [..]/alps/src/alps/ngs/detail/paramproxy.hpp on 78 in cast > [..]applications/dmft/qmc/hybridization(_ZNK4alps6detail10paramproxy4castIiEET_v+0x13d) > [0x42dbad] > [..]applications/dmft/qmc/hybridization(_ZN13hybridizationC1ERKN4alps6paramsEi+0x10b8) > [0x43f968] > [..]applications/dmft/qmc/hybridization(main+0x103) [0x41dfe3] > /lib64/libc.so.6(__libc_start_main+0xfd) [0x7ffff415fcdd] > [..]applications/dmft/qmc/hybridization() [0x41d409] > > I tested a previous version 2.1.1-r6670 with identical compilation process > on the same input files and it works flawlessly. > > There seem to be some significant changes in this version: There is a new > hybridization2 folder in applications/dmft/qmc/ and also the Documentation > inside this directory vanished. Do I have to change something in my > parm.h5 input file for this new version? > > Thanks for your help. > Best wishes, > Steffen > > > > ------------------------------ > > Message: 2 > Date: Wed, 21 Aug 2013 12:05:05 +0000 > From: Joseph Prentice <Joseph.Prentice(a)physics.ox.ac.uk> > To: "comp-phys-alps-users(a)lists.phys.ethz.ch" > <comp-phys-alps-users(a)lists.phys.ethz.ch> > Subject: [ALPS-users] Time propagation in TEBD stops prematurely > Message-ID: > <747861816F6A5C47BE50FCA07F5B2B4A0E25BB(a)EXCHNG16.physics.ox.ac.uk> > Content-Type: text/plain; charset="us-ascii" > > Hi all, > > I'm having a bit of trouble with getting a quench using the XXZ model to run in TEBD. I'm running the job on a Linux cluster, quenching Delta instantaneously from 3 to 6, and the program stops earlier than it should. I turned on the Verbose option, and it seems as though the imaginary time propagation works fine, but the real time propagation stops prematurely, with no error message. Does anyone have any idea why this has happened? It seems very odd. I will attach my Python script in case there is something wrong with it, and the output I obtained. > > Thanks very much, > > Joseph Prentice >

2 1

Re: [ALPS-users] Comp-phys-alps-users Digest, Vol 89, Issue 9
by Joseph Prentice 11 Sep '13

11 Sep '13

Hi Michael, sorry I've been slow on the reply, I didn't see this response. Thank you very much, this seems to have worked! My code had worked for lower Chi values, but when I increased the Chi value beyond a certain point, I experienced the problem I mentioned. Thank you for solving it! Joseph ------------------------------ Message: 2 Date: Wed, 21 Aug 2013 12:05:05 +0000 From: Joseph Prentice <Joseph.Prentice(a)physics.ox.ac.uk> To: "comp-phys-alps-users(a)lists.phys.ethz.ch" <comp-phys-alps-users(a)lists.phys.ethz.ch> Subject: [ALPS-users] Time propagation in TEBD stops prematurely Message-ID: <747861816F6A5C47BE50FCA07F5B2B4A0E25BB(a)EXCHNG16.physics.ox.ac.uk> Content-Type: text/plain; charset="us-ascii" Hi all, I'm having a bit of trouble with getting a quench using the XXZ model to run in TEBD. I'm running the job on a Linux cluster, quenching Delta instantaneously from 3 to 6, and the program stops earlier than it should. I turned on the Verbose option, and it seems as though the imaginary time propagation works fine, but the real time propagation stops prematurely, with no error message. Does anyone have any idea why this has happened? It seems very odd. I will attach my Python script in case there is something wrong with it, and the output I obtained. Thanks very much, Joseph Prentice

1 0

Re: [ALPS-users] Installing TEBD using MacPorts
by Joseph Prentice 11 Sep '13

11 Sep '13

Ah, OK. Thank you very much, I'll have a go with the binaries. Joseph ---------------------------------------------------------------------- Message: 1 Date: Tue, 10 Sep 2013 11:41:45 +0000 From: Joseph Prentice <Joseph.Prentice(a)physics.ox.ac.uk> To: "comp-phys-alps-users(a)lists.phys.ethz.ch" <comp-phys-alps-users(a)lists.phys.ethz.ch> Subject: [ALPS-users] Installing TEBD using MacPorts Message-ID: <747861816F6A5C47BE50FCA07F5B2B4A0E2C10(a)EXCHNG16.physics.ox.ac.uk> Content-Type: text/plain; charset="us-ascii" Hi all, I'm trying to get TEBD installed on an iMac running OSX 10.7.5 using MacPorts. The basic ALPS package installs fine, but TEBD is not installed with this, and I can't see any variants or so on to allow me to install the TEBD section of the code too. I presume that to create such a variant would mean editing the portfile in some way, but I wondered whether there was an easier method I could do myself in the Terminal. Does anyone know the best way to get TEBD installed using MacPorts? Thank you very much, Joseph Prentice ------------------------------ Message: 2 Date: Tue, 10 Sep 2013 14:14:39 +0200 From: Matthias Troyer <troyer(a)phys.ethz.ch> To: comp-phys-alps-users(a)lists.phys.ethz.ch Subject: Re: [ALPS-users] Installing TEBD using MacPorts Message-ID: <33FCA1CF-CD9A-490D-A651-EBB81CE9EB7D(a)phys.ethz.ch> Content-Type: text/plain; charset=us-ascii TEBD is a Fortran code and requires consistent installation of the Fortran HDF5 libraries. This is hard to do. If you need TEBD on the Mac I suggest that you download the binary Mac installer for ALPS Matthias On Sep 10, 2013, at 1:41 PM, Joseph Prentice <Joseph.Prentice(a)physics.ox.ac.uk> wrote: > Hi all, > > I'm trying to get TEBD installed on an iMac running OSX 10.7.5 using MacPorts. The basic ALPS package installs fine, but TEBD is not installed with this, and I can't see any variants or so on to allow me to install the TEBD section of the code too. I presume that to create such a variant would mean editing the portfile in some way, but I wondered whether there was an easier method I could do myself in the Terminal. Does anyone know the best way to get TEBD installed using MacPorts? > > Thank you very much, > > Joseph Prentice End of Comp-phys-alps-users Digest, Vol 90, Issue 4 ***************************************************

2 1

Installing TEBD using MacPorts
by Joseph Prentice 10 Sep '13

10 Sep '13

Hi all, I'm trying to get TEBD installed on an iMac running OSX 10.7.5 using MacPorts. The basic ALPS package installs fine, but TEBD is not installed with this, and I can't see any variants or so on to allow me to install the TEBD section of the code too. I presume that to create such a variant would mean editing the portfile in some way, but I wondered whether there was an easier method I could do myself in the Terminal. Does anyone know the best way to get TEBD installed using MacPorts? Thank you very much, Joseph Prentice

2 1

[ALPS-user] Simulate fermion Hubbard using dirloop_sse
by Tao Ying 08 Sep '13

08 Sep '13

Dear developers, I want to simulate the fermion Hubbard model using dirloop_sse, how can I write the parameter file? I did the following: LATTICE="square lattice"; L=8; W=8; MODEL="fermion Hubbard"; U = 5.0; V = 0.0; t = 1.0; T = 0.05; SWEEPS=2000; THERMALIZATION=5000; { mu=0.0; } Then an error occured: Cannot convert fermionic operator to a bosonic matrix. What's wrong with it? If I want to measure the structure factor or correlation function, do I need to add the following in the parameter file? REPRESENTATION="SSE"; MEASURE[Correlations]=true; MEASURE[Structure Factor]=true; MEASURE[Green Function]=true; Best, TaoYing -- Tao Ying, Ph.D candidate, Physics department Harbin Institute of Technology

2 4

Parallel ALPS (using python) jobs via SLURM
by Scott Garland 05 Sep '13

05 Sep '13

I'm attempting to execute tutorial3a.py in the mc-03-magnetization directory using MPI. I'm using a cluster running Ubuntu Server 12.04 LTS and using the SLURM queue manager. I have run it successfully in serial by submitting it to nodes via SLURM. I have also successfully run it in serial without SLURM. When running in serial, I find the following line of code inside the .py file. res = pyalps.runApplication('dirloop_sse',input_file,Tmin=5,writexml=True) According to the ALPS tutorial webpage, the simulation can be run in parallel by simply changing the line above to res = pyalps.runApplication('dirloop_sse',input_file,Tmin=5,writexml=True,MPI=4) All I have done is appended "MPI=4" as an argument. This generates the following output when run with or without SLURM: parsing task files ... Creating a new simulation: 1 on 1 processes Created run 1 locally All processes have been assigned NUMBER_OF_WORMS_PER_SWEEP: 39 Checking if Simulation 1 is finished: Finished Halted Simulation 1 WARNING: Unclosed tag: SIMULATION! boost::filesystem::rename: No such file or directory: "/home/sgarland/moreTutorials/alps/mc-03-magnetization/outputFolder/output.task1.out.run1.h5.bak", "/home/sgarland/moreTutorials/alps/mc-03-magnetization/outputFolder/output.task1.out.run1.h5" -------------------------------------------------------------------------- MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD with errorcode -2. NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes. You may or may not see output from other processes, depending on exactly when Open MPI kills them. Does anyone have any suggestions regarding this issue? I have tried both defining and not defining the number of cores for SLURM to allocate, but haven't had any luck. Thanks! Chad Garland

3 3

problem with ngs.py
by Borghi Giovanni 02 Sep '13

02 Sep '13

Dear Alps staff, I downloaded Alps in order to use the maximum entropy code for some analytic continuations, and wanted to start understanding it by using the script by Pruschke, found in a previous post, https://lists.phys.ethz.ch/pipermail/comp-phys-alps-users/2013/001978.html So after installing alps from source, using package alps-2.1.1-r6176-src-with-boost and following the instructions (I installed it on a linux machine with Ubuntu)... .. I performed the tutorial dmft-04-mott/tutorial4a.py, to produce a sample imaginary time Green's function data, and afterwards, from ipython, tried to follow some of the instructions of Pruschke's script, quoting from that script: "from pyalps.ngs import h5ar # hdf5 interface" this produced the error message: >/opt/alps/lib/pyalps/ngs.py in <module>() > 29 import types > 30 >---> 31 from pyngsparams_c import * > 32 params.__bases__ = (MutableMapping, ) + params.__bases__ > 33 ImportError: No module named pyngsparams_c And indeed I could not find any such module in the source. My question is: what is going on? Did I miss to install some pre-requisite? Note that the tutorial simulation for the hubbard model works fine till the end, same thing can be said for all tests which I ran before installing the code in its /opt/alps path. Thank you in advance for your help, Giovanni

2 1

Re: [ALPS-users] Alps 2.1.2 r6963 hybridization expansion: No parameter available
by backes＠th.physik.uni-frankfurt.de 22 Aug '13

22 Aug '13

Thanks for looking into it! I tried the newest SVN version 6991 and except for one thing everything went smoothly. My last problem is resolved now. I encountered the following error on compilation: During cmake, I got: CMake Error at applications/qmc/dwa/CMakeLists.txt:27 (install): install TARGETS given target "worldlines_c.so" which does not exist in this directory. I commented out the lines mentioned above, since I don't need it and compilation continued successfully. best wishes, Steffen > Steffen, > > it seems that the hybridization3 code was not deemed important enough to > make it into the 2.1 branch. We're working on fixing this (thanks for > telling!!!). For now, the svn code should be stable and have all the > things you need. Can you confirm? > > Cheers, > Emanuel > > On Aug 21, 2013, at 12:45 PM, backes(a)th.physik.uni-frankfurt.de wrote: > >> Hi Emanuel, >> >> thanks for the quick answer! Here is the parameter file I've been using >> for both versions of the code. alps-2.1.2-r6963 gives the error, >> alps-2.1.1-r6670 runs fine. >> But I will try the newest version from the SVN repositories and will >> report the results. >> >> SEED = 0 >> THERMALIZATION = 1000 >> SWEEPS = 10000000 >> MAX_TIME = 60 >> BETA = 40.0 >> N_MEAS = 50 >> N_HISTOGRAM_ORDERS = 50 >> N_ORBITALS = 10 >> U_MATRIX = "umatrix.dat" >> MU_VECTOR = "mu_vector.dat" >> DELTA = "delta_0.dat" >> N_TAU = 1000 >> TEXT_OUTPUT = 1 >> VERBOSE = 1 >> OUTPUT_PERIOD = 100000 >> MEASURE_freq = 0 >> MEASURE_legendre = 1 >> N_LEGENDRE = 50 >> N_MATSUBARA = 500 >> >> best wishes, >> Steffen >> >>> Hi Steffen, >>> >>> could you please send me the new and old parameter file? There have >>> been >>> quite a few changes since December and I'll need to figure out which of >>> these caused the problem... >>> >>> Emanuel >>> >>> On Aug 21, 2013, at 7:51 AM, backes(a)th.physik.uni-frankfurt.de wrote: >>> >>>> Dear all, >>>> >>>> I want to use the newest version of the ALPS Hybridization Expansion >>>> Impurity Solver from the nightly build 2.1.2-r6963. >>>> Compilation is successful, but when running the solver with a simple >>>> test >>>> case I get the following Error (I shortened the paths for >>>> readability): >>>> >>>>> hybridization parm.h5 >>>> >>>> terminate called after throwing an instance of 'std::runtime_error' >>>> what(): No parameter available >>>> In [..]/alps/src/alps/ngs/detail/paramproxy.hpp on 78 in cast >>>> [..]applications/dmft/qmc/hybridization(_ZNK4alps6detail10paramproxy4castIiEET_v+0x13d) >>>> [0x42dbad] >>>> [..]applications/dmft/qmc/hybridization(_ZN13hybridizationC1ERKN4alps6paramsEi+0x10b8) >>>> [0x43f968] >>>> [..]applications/dmft/qmc/hybridization(main+0x103) [0x41dfe3] >>>> /lib64/libc.so.6(__libc_start_main+0xfd) [0x7ffff415fcdd] >>>> [..]applications/dmft/qmc/hybridization() [0x41d409] >>>> >>>> I tested a previous version 2.1.1-r6670 with identical compilation >>>> process >>>> on the same input files and it works flawlessly. >>>> >>>> There seem to be some significant changes in this version: There is a >>>> new >>>> hybridization2 folder in applications/dmft/qmc/ and also the >>>> Documentation >>>> inside this directory vanished. Do I have to change something in my >>>> parm.h5 input file for this new version? >>>> >>>> Thanks for your help. >>>> Best wishes, >>>> Steffen >>>> >>> >>> >> > >

1 0

Alps 2.1.2 r6963 hybridization expansion: No parameter available
by backes＠th.physik.uni-frankfurt.de 21 Aug '13

21 Aug '13

Dear all, I want to use the newest version of the ALPS Hybridization Expansion Impurity Solver from the nightly build 2.1.2-r6963. Compilation is successful, but when running the solver with a simple test case I get the following Error (I shortened the paths for readability): > hybridization parm.h5 terminate called after throwing an instance of 'std::runtime_error' what(): No parameter available In [..]/alps/src/alps/ngs/detail/paramproxy.hpp on 78 in cast [..]applications/dmft/qmc/hybridization(_ZNK4alps6detail10paramproxy4castIiEET_v+0x13d) [0x42dbad] [..]applications/dmft/qmc/hybridization(_ZN13hybridizationC1ERKN4alps6paramsEi+0x10b8) [0x43f968] [..]applications/dmft/qmc/hybridization(main+0x103) [0x41dfe3] /lib64/libc.so.6(__libc_start_main+0xfd) [0x7ffff415fcdd] [..]applications/dmft/qmc/hybridization() [0x41d409] I tested a previous version 2.1.1-r6670 with identical compilation process on the same input files and it works flawlessly. There seem to be some significant changes in this version: There is a new hybridization2 folder in applications/dmft/qmc/ and also the Documentation inside this directory vanished. Do I have to change something in my parm.h5 input file for this new version? Thanks for your help. Best wishes, Steffen

3 3