triggering_on_cluster.submit_multibanks#

Back to Module page

Summary#

Submits multibanks to the cluster

Signature#

def submit_multibanks(tbp_to_use, files_to_submit = None, bank_ids = BBH_KEYS, cluster = 'typhon', observing_run = 'O3a', test_few = None, output_dirs = None, submit = True, save_hole_correction = True, preserve_max_snr = params.DEF_PRESERVE_MAX_SNR, fmax = params.FMAX_OVERLAP, n_cores = None, n_hours_limit = 24, fftlog2size = DEFAULT_FFTLOG2SIZE, njobchunks = 1, trim_empty = False, queue_name = None, env_command = None, use_HM = False, mem_per_cpu = 4, exclude_nodes = False)

Input variables#
Name	Type	Default	Description
`tbp_to_use`			Template bank parameter object to use
`files_to_submit`		None	If desired, list of files to run
`bank_ids`		BBH_KEYS	List of bank_ids, keys to dictionaries in template_bank_params….py
`cluster`		‘typhon’	Name of cluster we are submitting jobs to
`observing_run`		‘O3a’	If files_to_submit is None, we use the name of observing run (can be “O1”, “O1new”, “O2”, “O3a”, or “O3b”)
`test_few`		None	If desired, number of files to submit as a test
`output_dirs`		None	Nested lists of output directories as n_bank_ids x n_subbanks. One can also specify list of particular sub-banks if needed, note that dirnames should end with subbank #. If None, we will generate the names in create_trig_dir_name()
`submit`		True	Flag whether to submit to the cluster
`save_hole_correction`		True	Flag whether to save hole correction, exposed here to rerun files in run “O1”
`preserve_max_snr`		params.DEF_PRESERVE_MAX_SNR	Exposed here since it changed between “O1” and subsequent runs
`fmax`		params.FMAX_OVERLAP	Exposed here since it changes between BBH-0 and the rest of the banks
`n_cores`		None	Number of cores to use for each file, we use sensible values if not provided
`n_hours_limit`		24	Number of hours to limit the job to. If n_hours_limit<=24, we may get into a faster queue in the cluster
`fftlog2size`		DEFAULT_FFTLOG2SIZE	log_2(fftsize) to use for submission
`njobchunks`		1	Number of chunks to split the job into, useful if we want to save trigger files in between and restart jobs (useful when n_hours_limit<=24)
`trim_empty`		False	Trim apparently empty files
`queue_name`		None	name of queue (used only in WEXAC submission)
`env_command`		None	If required, pass the command to activate the conda environment If None, will infer from the username if it’s in utils
`use_HM`		False	Boolean flag indicating whether we want to use 33 and 44 modes alongside 22
`mem_per_cpu`		4	Memory per CPU in GB. 4GB/core is the default in Typhon
`exclude_nodes`		False	Flag to exclude nodes from the job submission (if other people start complaining about cluster use)

Output variables#

Return annotation	Docstring type	Description
`None`

Docstring#

Submits multibanks to the cluster
:param tbp_to_use: Template bank parameter object to use
:param files_to_submit: If desired, list of files to run
:param bank_ids:
    List of bank_ids, keys to dictionaries in template_bank_params....py
:param cluster: Name of cluster we are submitting jobs to
:param observing_run: If files_to_submit is None,
    we use the name of observing run (can be "O1", "O1new", "O2",
                      "O3a", or "O3b")
:param test_few: If desired, number of files to submit as a test
:param output_dirs:
    Nested lists of output directories
    as n_bank_ids x n_subbanks.
    One can also specify list of particular sub-banks if needed,
    note that dirnames should end with subbank #.
    If None, we will generate the names in create_trig_dir_name()
:param submit: Flag whether to submit to the cluster
:param save_hole_correction:
    Flag whether to save hole correction, exposed here to rerun files
    in run "O1"
:param preserve_max_snr:
    Exposed here since it changed between "O1" and subsequent runs
:param fmax:
    Exposed here since it changes between BBH-0 and the rest of the banks
:param n_cores:
    Number of cores to use for each file, we use sensible values if
    not provided
:param n_hours_limit:
    Number of hours to limit the job to.
    If n_hours_limit<=24, we may get into a faster queue in the cluster
:param fftlog2size: log_2(fftsize) to use for submission
:param njobchunks:
    Number of chunks to split the job into, useful if we want to save
    trigger files in between and restart jobs
    (useful when n_hours_limit<=24)
:param trim_empty: Trim apparently empty files
:param queue_name: name of queue (used only in WEXAC submission)
:param env_command:
        If required, pass the command to activate the conda environment
        If None, will infer from the username if it's in utils
:param use_HM:
        Boolean flag indicating whether we want to use 33 and 44 modes alongside 22
:param mem_per_cpu: Memory per CPU in GB. 4GB/core is the default in Typhon
:param exclude_nodes: Flag to exclude nodes from the job submission
                    (if other people start complaining about cluster use)
:return: