stilt.execution.worker.run_receptors#

stilt.execution.worker.run_receptors(project, receptor_ids, *, compute_root, execution=None, skip_existing=True)[source]#

Run a list of receptors, in this process or in a process pool.

The variants of a batched transport model run first, one model call for all the receptors that need their particles. Then each receptor runs its other variants. Pool workers open the project again from its directory, with the same output directory. Ctrl-C, or a SIGTERM such as Slurm preemption or a stopped pod, stops the batch: the transport model runs in progress are ended, the workers exit, and this logs how many receptors had finished and returns. What finished is in the output directory, with a failure record for each simulation that failed; a run that was cut off leaves only its log line, so it is interrupted in stilt.Project.status().

Parameters:

  • project (Project) – Project the receptors belong to.

  • receptor_ids (list[str]) – Receptors to run.

  • compute_root (Path) – Directory the workdirs are made in, a new one for each run, as resolve_compute_root() returns it.

  • execution (ExecutionConfig | None) – Execution settings: cpus is the number of worker processes (1 runs in this process), and timeout and keep_workdir apply to each run. Defaults to the project’s.

  • skip_existing (bool) – Keep particles and footprints that already exist.

Return type:

None