-
Notifications
You must be signed in to change notification settings - Fork 189
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
lj-demo.py is broken #4347
Comments
I will take a look. |
@davidbbeyer any update? |
Sorry, I haven't fixed the overflow issue yet, I will do it in the next days. |
kodiakhq bot
added a commit
that referenced
this issue
Mar 22, 2022
Description of changes: - improve exception handling in MPI code (partial fix for #4219) - several runtime errors used to be queued for far too long and would only surface during integration - runtime errors from non-bonded interactions instantiated with a cutoff too large now immediately raise an exception, from which the user needs to recover by either reducing the cutoff or disabling the interaction - runtime errors from virtual sites relative applied to particles too far away now immediately raise an exception, from which the user needs to recover by increasing the minimum global cutoff - all automatically-generated script interface methods used to call `handle_errors("")` with an empty string; now it's called with the method name to help identify which feature threw the exception - virtual sites fatal errors are now runtime errors - quaternion fatal errors are now runtime errors - invalid particle ids fatal errors are now value errors - fix bugs - virtual sites can no longer relate to themselves - particles can no longer be created with a negative id - LB node property `boundary` no longer returns a random integer when `LB_BOUNDARIES` is not compiled in - improve code coverage - common exceptions are now properly covered by tests (e.g. error messages from the cell system and integrator) - rotation code is now unit tested - h5md exceptions are now tested - fix regressions in the testsuite - checkpoint tests were only running on 1 MPI core - the LB VTK had a wall boundary misplaced - virtual sites relative test cases can now run in any order - fix regression in the OpenGL visualizer - activating the `LB_draw_boundaries` option without any other `LB_draw_*` option no longer throws an exception - rescaling the box size in `lj-demo.py` now properly rescales the particle coordinates (partial fix for #4347)
kodiakhq bot
added a commit
that referenced
this issue
Apr 4, 2022
Description of changes: - API change: rename `espressomd.reaction_ensemble` to `espressomd.reaction_methods` - feature configuration: remove obsolete `EXPERIMENTAL_FEATURES` - script interface: - use shared pointers instead of `new` expressions in the script interface - make type conversion error messages less ambiguous by putting quotes around C++ types - require EK node grid indices to be exactly 3 integers (2 indices would skip the exception and return `None`) - python interface: - replace `from X import Y` by `import X` and `X.Y` in samples and python interface - use `tqdm`, `numpy` and f-strings more often in the samples - remove `lj-demo.py` (fixes #4347) - remove mentions to non-existent functions in the user guide
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The physics in
samples/lj-demo.py
is broken since ESPResSo 4.0. The script is also completely untested, and has diverged, e.g. the NpT option doesn't work because the integrator function signature changed in 4.1. It's possible to fix the sample without a MIDI board.Tasks:
testsuite/scripts/samples/test_visualization_npt.py
can be used as template, should probably work with the following patch:patch (click to unroll)
The text was updated successfully, but these errors were encountered: