Re: [gmx-users] trjcat skipping frames
Jared James Thompson wrote: Hi All, I'm having a problem with trjcat skipping frames. I saw a similar complaint seemingly handled by Berk Hess in 2008. I wanted to know if his bug fix has been incorporated into the current gromacs version. Can you provide a link to this information? What version of Gromacs are you using? Presumably, if you're using the latest version, bugs from 2 years ago should be fixed. I have a number of short .trj files of equal length and time spacing that I want to cat together to form a single .trj file. I have a method for doing this using trjcat and trjconv. It has worked normally in the past. The method is as follows. Set the start time of each trajectory to be the next putative frame immediately following the previous trajectory, and also set the timestep for each frame correctly, so for .trj of 10 frames long: trjconv -f frame_traj_2.trr -o ftc_2.trr -t0 20 -timestep 1.0 trjconv -f frame_traj_3.trr -o ftc_3.trr -t0 30 -timestep 1.0 etc. Once all the frame_traj_x files are together, then use trjcat to cat them. trjcat -f ftc* -o CombinedTraj.trj However, this time that I am using it, trjcat is producing the following response, note the timestep shown in the File Roster and that ftc_5 is treated normally for some reason: FileStart time Time step - ftc_1.trr0.000 ps1.000 ps ftc_2.trr 20.000 ps0.000 ps ftc_3.trr 40.000 ps0.000 ps Clearly something has broken down on a very fundamental level. You set ftc_3.trr to have a start time of 30 ps above, but it has not worked. Some questions that may help diagnose your issue: 1. What is the spacing of frames in your original .trr files? 2. What does gmxcheck tell you about each of your ftc_*.trr files? 3. Does passing -t0 and -timestep to trjconv in separate steps make any difference? -Justin -- Justin A. Lemkul Ph.D. Candidate ICTAS Doctoral Scholar MILES-IGERT Trainee Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
[gmx-users] trjcat skipping frames
Hi All, I'm having a problem with trjcat skipping frames. I saw a similar complaint seemingly handled by Berk Hess in 2008. I wanted to know if his bug fix has been incorporated into the current gromacs version. I have a number of short .trj files of equal length and time spacing that I want to cat together to form a single .trj file. I have a method for doing this using trjcat and trjconv. It has worked normally in the past. The method is as follows. Set the start time of each trajectory to be the next putative frame immediately following the previous trajectory, and also set the timestep for each frame correctly, so for .trj of 10 frames long: trjconv -f frame_traj_2.trr -o ftc_2.trr -t0 20 -timestep 1.0 trjconv -f frame_traj_3.trr -o ftc_3.trr -t0 30 -timestep 1.0 etc. Once all the frame_traj_x files are together, then use trjcat to cat them. trjcat -f ftc* -o CombinedTraj.trj However, this time that I am using it, trjcat is producing the following response, note the timestep shown in the File Roster and that ftc_5 is treated normally for some reason: FileStart time Time step - ftc_1.trr0.000 ps1.000 ps ftc_2.trr 20.000 ps0.000 ps ftc_3.trr 40.000 ps0.000 ps ftc_4.trr 60.000 ps0.000 ps ftc_5.trr 80.000 ps1.000 ps Back Off! I just backed up CombinedTraj.trr to ./#CombinedTraj.trr.3# Reading frame 0 time0.000 Continue writing frames from ftc_1.trr t=0 ps, frame=0 Last frame 20 time 20.000-> frame 10 time 10.000 ps Reading frame 0 time 20.000 lasttime 19 Continue writing frames from ftc_2.trr t=20 ps, frame=20 Last frame 0 time 20.000 WARNING: Frames around t=20.00 ps have a different spacing than the rest, might be a gap or overlap that couldn't be corrected automatically. Reading frame 0 time 40.000 lasttime 20 Continue writing frames from ftc_3.trr t=40 ps, frame=21 Last frame 0 time 40.000 WARNING: Frames around t=40.00 ps have a different spacing than the rest, might be a gap or overlap that couldn't be corrected automatically. Reading frame 0 time 60.000 lasttime 40 Continue writing frames from ftc_4.trr t=60 ps, frame=22 Last frame 0 time 60.000 WARNING: Frames around t=60.00 ps have a different spacing than the rest, might be a gap or overlap that couldn't be corrected automatically. Reading frame 0 time 80.000 lasttime 60 Continue writing frames from ftc_5.trr t=80 ps, frame=23 Last frame 20 time 100.000-> frame 40 time 97.000 ps Last frame written was 43, time 100.00 ps Any help would be appreciated. Best, Jared Thompson -- Jared James Thompson Department of Medicinal Chemistry and Molecular Pharmacology Laboratory for Computational Drug Design and Biology RHPH 504C Heine Pharmacy Building 575 Stadium Mall Drive West Lafayette, IN 47907-2091 -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] mpi installation problems
Hi Ryan, you need to use mpiicc instead of mpicc (should be present also in /opt/intel/mpi/3.1/bin64/). Carsten On Jul 28, 2010, at 11:32 PM, Ryan S Davis (rsdavis1) wrote: > I am trying to install gromacs with MPI enabled on a cluster but it seems > like fftw is giving me trouble. > > First, I compile everything without MPI just fine. FFTW is already on the > cluster. I export pertinent variables... > > export CPPFLAGS=-I/opt/fftw/3.2.1/include > export LDFLAGS=-L/opt/fftw/3.2.1/lib > export CC=/opt/intel/Compiler/11.1/069/bin/intel64/icc > export F77=/opt/intel/Compiler/11.1/069/bin/intel64/ifort > > I have been using gromacs without MPI for over a month without any issues. > Now, I need to recompile mdrun with MPI. So I use... > > export MPICC=/opt/intel/mpi/3.1/bin64/mpicc > > > in addition to all the previous exports. Then, I issue the command... > >> ./configure --prefix=/home/username/gromacs --enable-mpi >> --program-suffix=_mpi --disable-nice --enable-double > > > which runs fine. Then I issue... > >> make mdrun > > and I end up with a bunch of undefined references. I am copying some of the > output here. Also, I do not know if FFTW is compiled with single/double > because of unresponsive admins, but I have tried compiling gromacs with both > options and still got errors. Also, from a previous post, I read the idea to > install FFTW with shared libraries enabled, which I tried but just ended up > with other errors. Thanks in advance for any help. I have been at this for > weeks. > > By the way, this might not be critical, but here is some system info > # 2 Nodes Share: >* 1x 1702 Altus Server (1U, contains 2 Independent Motherboards) > # 1x AMD Dual Opteron 2380(Shanghai) 2.5GHz Quad Core (8 Cores) > # 250GB SATA2 Hard Drive > > *output from "make mdrun" ** > > /opt/fftw/3.2.1/lib/libfftw3.a(vrank3-transpose.o):vrank3-transpose.c:(.text+0x649): > more undefined references to `_inte > l_fast_memcpy' follow > /opt/fftw/3.2.1/lib/libfftw3.a(vrank3-transpose.o): In function > `apply_toms513': > vrank3-transpose.c:(.text+0xe3e): undefined reference to `_intel_fast_memset' > vrank3-transpose.c:(.text+0xef4): undefined reference to `_intel_fast_memcpy' > vrank3-transpose.c:(.text+0xf0e): undefined reference to `_intel_fast_memcpy' > vrank3-transpose.c:(.text+0x106d): undefined reference to `_intel_fast_memcpy' > vrank3-transpose.c:(.text+0x108a): undefined reference to `_intel_fast_memcpy' > vrank3-transpose.c:(.text+0x110d): undefined reference to `_intel_fast_memcpy' > /opt/fftw/3.2.1/lib/libfftw3.a(vrank3-transpose.o):vrank3-transpose.c:(.text+0x1122): > more undefined references to `_int > el_fast_memcpy' follow > /opt/fftw/3.2.1/lib/libfftw3.a(dht-rader.o): In function `apply': > dht-rader.c:(.text+0x4a8): undefined reference to `_intel_fast_memset' > /opt/fftw/3.2.1/lib/libfftw3.a(dht-rader.o): In function `awake': > dht-rader.c:(.text+0xaab): undefined reference to `_intel_fast_memset' > /opt/fftw/3.2.1/lib/libfftw3.a(cpy1d.o): In function `fftw_cpy1d': > cpy1d.c:(.text+0x268): undefined reference to `_intel_fast_memcpy' > /opt/fftw/3.2.1/lib/libfftw3.a(cpy2d.o): In function `dotile_buf': > cpy2d.c:(.text+0x6a7): undefined reference to `_intel_fast_memcpy' > cpy2d.c:(.text+0x129f): undefined reference to `_intel_fast_memcpy' > cpy2d.c:(.text+0x1870): undefined reference to `_intel_fast_memcpy' > cpy2d.c:(.text+0x1902): undefined reference to `_intel_fast_memcpy' > /opt/fftw/3.2.1/lib/libfftw3.a(cpy2d.o):cpy2d.c:(.text+0x22dd): more > undefined references to `_intel_fast_memcpy' follow > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Unpack_external' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Win_get_errhandler' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to `PMPI_Win_start' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to `PMPI_Win_test' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Comm_spawn' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Request_get_status' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `MPIR_Err_return_comm' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to `PMPI_Type_dup' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Win_get_group' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Comm_spawn_multiple' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to `PMPI_Free_mem' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `MPIR_Grequest_set_lang_f77' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Win_create_keyval' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `b_use_gettimeofday' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_Type_get_attr' > /opt/intel/impi/3.1/lib64/libmpigf.so: undefined reference to > `PMPI_