Search found 2564 matches
- Thu Mar 12, 2009 2:57 pm
- Forum: SimTK Core Toolset
- Topic: Inverse Dynamics
- Replies: 17
- Views: 2833
RE: Inverse Dynamics
<t>Hi,<br/> <br/> If you're getting a linking error, the problem is probably in the way you're compiling or running, not in the code. Does the error occur at compile time or run time? You need to make sure you have all the proper libraries in your compile/run path. Also, are you statically or dynami...
- Wed Mar 11, 2009 11:47 am
- Forum: OpenMM
- Topic: Integrator->step question
- Replies: 1
- Views: 361
RE: Integrator->step question
<r>After you create your OpenMMContext, call<br/> <br/> context.setPositions(positions);<br/> <br/> to tell it the positions of the atoms. To get the forces, you can then call<br/> <br/> State state = context.getState(State::Forces);<br/> <br/> The force on atom i is given by state.getForces()<I><s>...
- Wed Mar 11, 2009 11:43 am
- Forum: OpenMM
- Topic: Installation problems
- Replies: 14
- Views: 1747
RE: Installation problems
<t>Oh dear, I'm afraid you've made things much more difficult for yourself than they needed to be! There's no need to edit any CMakeList.txt files. The easy way to use CMake is to run the "ccmake" command. It provides a UI that lists all the available options and lets you configure them. One of the ...
- Tue Mar 10, 2009 5:42 pm
- Forum: OpenMM
- Topic: Installation problems
- Replies: 14
- Views: 1747
RE: Installation problems
<t>The .a files are only needed if you want to statically link OpenMM into your own code. As long as you link to it dynamically, you want the .so files.<br/> <br/> Make sure that the openmm/lib directory is in the compiler's library path (use the -L flag with gcc), and also that you specify -lOpenMM...
- Tue Mar 10, 2009 2:00 pm
- Forum: OpenMM
- Topic: Installation problems
- Replies: 14
- Views: 1747
RE: Installation problems
You need to include the openmmapi/include folder from the source code in your include path. With gcc, you do that using the -I flag.
Peter
Peter
- Tue Mar 10, 2009 12:37 pm
- Forum: OpenMM
- Topic: Installation problems
- Replies: 14
- Views: 1747
RE: Installation problems
Correct. You'll need to include the header files for all the classes you use (OpenMMContext.h, System.h, etc.).
We hope to make a lot of these things easier in the next version.
Peter
We hope to make a lot of these things easier in the next version.
Peter
- Tue Mar 10, 2009 11:36 am
- Forum: OpenMM
- Topic: Installation problems
- Replies: 14
- Views: 1747
RE: Installation problems
The CUDA code won't work in emulation mode, but OpenMM includes a reference implementation written in C/C++. It's slow, but it works fine for just trying out the API and seeing how it works.
Peter
Peter
- Tue Mar 10, 2009 11:30 am
- Forum: OpenMM
- Topic: Installation problems
- Replies: 14
- Views: 1747
RE: Installation problems
Hi Siddharth,
No, the message means exactly what it says: CMake 2.6 is required.
Have you tried the precompiled binaries? You may not need to recompile it from source.
Peter
No, the message means exactly what it says: CMake 2.6 is required.
Have you tried the precompiled binaries? You may not need to recompile it from source.
Peter
- Thu Mar 05, 2009 6:52 pm
- Forum: OpenMM
- Topic: Problem with Mac Version
- Replies: 1
- Views: 308
RE: Problem with Mac Version
<r>Hi Justin,<br/> <br/> The message means that your GPU doesn't have enough memory to simulate the system. How many atoms does it include? And what GPU do you have? High end ones will typically have 1 GB or more of memory, while low end ones might only have 256 MB.<br/> <br/> I've done a lot of wor...
- Tue Feb 24, 2009 6:52 pm
- Forum: OpenMM
- Topic: Is 32-bit floating point good enough?
- Replies: 1
- Views: 411
RE: Is 32-bit floating point good enough?
<t>I think there are a few different questions mixed together there.<br/> <br/> The precompiled binaries we distribute are all 32 bit. That refers to the size of pointers, not to the size of floating point numbers. Unless you want to access very large amounts of memory, 32 bit binaries are preferabl...