Re: [ESPResSo-users] Error in python code checks

[Rudolf Weeber]

Hi Michael,
On Wed, Mar 28, 2018 at 08:55:09PM +0200, Michael Klatt wrote:
> -------------------------------------------------------------------------
> 
> (1) When I compiled the latest release (3.3.1), "import espressomd" 
> causes no problems, but "print(espressomd.features())" throws an error:
> Traceback (most recent call last):
>    File "<stdin>", line 1, in <module>
> AttributeError: 'module' object has no attribute 'features'
> 
> -------------------------------------------------------------------------
The stuff from the tcl branch is outdated and will not be pursued further. 
Please use the Python branch from the github page.


> (2) Last week, I have cloned the python branch of ESPResSo and followed 
> the instructions for installation from the ESPResSo documentation.
> 
> "make install" was completed without errors.
> However, "make check" reported a problem:
> undefined reference to `boost::mpi::communicator::operator int() const'
> 
> When I import espressomd in "pypresso", the following error is reported:
> libcluster_analysis.so.4: undefined symbol: 
> _ZNK5boost3mpi12communicatorcviEv
> 
> -------------------------------------------------------------------------
First of all, did you build in an empty build directory (i.e., not the oone you 
used for the 3.3 code before)?
If not, please try that.

Otherwise, maybe the boost mpi library is not installed or was not detected.
On Ubuntu, e.g., you get it via
sudo apt-get install libboost-mpi-dev
If you have the boost-mpi package installed, please post the output of cmake 
(again, starting from an empty build dir).

We have (as part of our automatic testing) Docker build images and scripts for 
several Linux distros. Maybe also for yours
github.com/espressomd/docker/tree/master/docker


If that doesn't solve it, on what platform are you?
In many cases, it's best to get started on a Desktop machine or a laptop, where 
you can install software freely.
On a compute cluster, things tend to be a bit more involved.
On BwUniCluster, e.g., you have to compile your own version of boost.

With regards to minimize_energy:

        system.integrator.set_steepest_descent(
            f_max=0, gamma=0.1, max_displacement=0.1)
        system.integrator.run(20)
        
        # Back to velocity Verlet
        system.integrator.set_vv()



Hope that helps!
Regards, Rudolf


--
Dr. Rudolf Weeber
Institute for Computational Physics
Universit??t Stuttgart
Allmandring 3
70569 Stuttgart
Germany
Phone: +49(0)711/685-67717
Email: address@hidden
http://www.icp.uni-stuttgart.de/~icp/Rudolf_Weeber