-
Notifications
You must be signed in to change notification settings - Fork 34
Open
Description
In the thermal code, the stack buffer is separate from the walkers buffer for MPI. The former should be concatenated to the latter such that only the walkers buffer is passed around in the population control algorithms. See for instance:
| stack_buff = walkers.stack[iw].get_buffer() |
| walkers.stack[iw].set_buffer(buff[walkers.buff_size:]) |
| buffer = numpy.concatenate((walkers.walker_buffer, walkers.stack[0].stack_buffer)) |
| buffer = numpy.concatenate((walkers.walker_buffer, walkers.stack[0].stack_buffer)) |
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels