Re: [ESPResSo-users] Subtracted LJ bond

[Clemens Jochum]

Hi Florian,

I see that the fix is already merged, thanks for the quick fix!
Unfortunately, I now have a different problem: When I try to run my MWE I get the following error message:

Traceback (most recent call last):
  File "/home/jochum/phd/subt_lj_test.py", line 23, in <module>
    subtLJbond = espressomd.interactions.Subt_Lj(k = 1., r = 100.)
  File "espressomd/interactions.pyx", line 1405, in espressomd.interactions.Subt_Lj.__init__ (/home/jochum/build/pypresso/src/python/espressomd/interactions.cpp:13560)
  File "espressomd/interactions.pyx", line 705, in espressomd.interactions.BondedInteraction.__init__ (/home/jochum/build/pypresso/src/python/espressomd/interactions.cpp:7984)
  File "espressomd/interactions.pyx", line 741, in espressomd.interactions.BondedInteraction.params.__set__ (/home/jochum/build/pypresso/src/python/espressomd/interactions.cpp:8499)
TypeError: cannot concatenate 'str' and 'method-wrapper' objects

It seems like a minor error, which should be easy to fix. But I always get lost trying to find the cause in the code.

Best,
Clemens

On 24.10.2017 17:18, Florian Weik wrote:

Hi Clemens,

this was indeed incorrect. It is fix in https://github.com/espressomd/espresso/pull/1564, 

which is currently under review and will be merged soon.

Cheers,

Florian

On Tue, Oct 24, 2017 at 12:12 PM Clemens Jochum <address@hidden> wrote:

Dear all,

I'm trying to use the subtracted LJ bond [1] and it is not working
correctly when I'm using LJ with offset. I'm using the current
development version of espresso.

I attached a MWE of my system without thermostat, where the conformation
is such that with subtracted LJ bond between all particles there should
be no forces present. Without offset and the subtracted LJ bond, the
forces are 0 as expected. But with offset and subtracted LJ bond the
forces are NAN or if I don't update the forces and try to integrate some
bond is immediately broken.

I think either the subtracted LJ-bond ignores the offset and calculates
the wrong (very large) value orthe subtracted LJ-bond does not calculate
at all below the offset and just reads in some NAN value.

Is there a way to fix this?

Best,
Clemens

[1] http://espressomd.org/html/doc/inter.html#subtracted-lennard-jones-bond