Skip to content
slurmtools

generate_nmm_config Reference

Using the model_number argument

This argument is a string of the model number. The default is NULL and this information is pulled from the .mod argument

Using the files_to_track argument

This argument sets which NONMEM files to watch for changes. Default is NULL which will cause nmm to track c(".ext", ".lst", ".grd") files. You can currently set to any NONMEM output file

Using the tmp_dir argument

This argument is not currently used by nmm, but will be in the future

Using the watched_dir argument

watched_dir = file.path(“model”, “nonmem”) This argument sets where nmm will look for model output folders to monitor NONMEM output files, The default will set this to here::here()/model/nonmem/<model number>, but this is specific to our internal A2-Ai directory structure

Using the output_dir argument

output_dir = file.path(watched_dir, “in_progress”) This argument sets where nmm will save the log file and output files it creates. The default is `here::here()/model/nonmem/in_progress’

Using the bbi_config_path argument

bbi_config_path = file.path(“model”, “nonmem”, “bbi.yaml”) This argument sets where the bbi config file is. The default is options('slurmtools.bbi_config_path)

Using the overwrite argument

This sets whether to overwrite a previously run model

Using the poll_duration argument

This argument sets how long before nmm polls the watched_dir for updates to the files. Default is NULL which will cause nmm to poll every second. This argument must be an integer > 1

Using the level argument

This sets the logging level for nmm options are Trace, Debug, Info, Warn, Fatal default is Info

Using the threads argument

The number of cpus to use for the model. If you want parallel processing set this > 1

Using the alerter_opts argument

A named list that accepts

  • alerter: string, path to binary to use for alerting
  • command: string, command to use with alerter
  • message_flag: string, flag to use for defining message, default is m
  • use_stdout: boolean, whether to capture standard out after initial call and use in subsequent calls
  • args: named list of additional arguments to be used with the alerter
alerter_opts = list(
  alerter = "path/to/alerter/binary",
  command = "notify_command",
  message_flag = "s",
  use_stdout = TRUE,
  args = list(
    email = "your_email@example.com",
    config = "path/to/alerter/config.file"
  )
)