Description Usage Arguments Details See Also Examples
This generic function is used to submit a Rscript job to a Slurm cluster. Current support is for Singularity and Conda environments.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | submit(x, ...)
## Default S3 method:
submit(
job.name,
script.file = "",
cluster = "",
slurm.args = "",
script.args = "",
conda = "",
singularity = "docker://rocker/r-base:latest",
r.bin.args = "",
target.dir = "~",
r.data = "",
local.data = "",
profile = JobProfile(),
save.profile.name = "",
VERBOSE = FALSE,
save.sub = FALSE
)
## S3 method for class 'Submission'
submit(sub, job.name = "R-slurmSubmit", save.sub = FALSE, VERBOSE = FALSE)
|
x |
A Submission class object, or a string containing a name for a new job submission. |
... |
Other arguments |
job.name |
Name of job to be submitted. REQUIRED for default method. |
script.file |
Relative path to the script |
cluster |
Address of cluster you want to submit to. |
slurm.args |
String containing all |
script.args |
String containing all arguments to the submitted Rscript to be executed remotely on the cluster. |
conda |
String containing the name of a valid conda environment on the cluster to be used to execute R. |
singularity |
String path or URI of a Singularity container to execute Rscript inside of. |
r.bin.args |
String containing all arguments to pass to the |
target.dir |
Path to directory to create .slurmsubmit/ directory. |
r.data |
A character vector containing the name(s) of valid R data objects in the search path. This data will be made
available locally on the cluster in the same directory as the script under the file handle |
local.data |
String containing a list of paths to one or more local file(s). WARNING: Large files may take a long time to sync. |
profile |
A JobProfile type object. |
save.profile.name |
If specified, will save the generated profile object under the specified name in .Globalenv. |
VERBOSE |
Show more details. |
save.sub |
If TRUE will save this job submission as a Submission type object under |
sub |
A Submission type object |
It is suggested that you use clean.remote() after your job has completed to clean the remote address of the .slurmSubmit/ hidden directory.
Submission, JobProfile, clean.remote
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 | # Naive method
submit("example", cluster="login.sumner.jax.org", slurm.args="-q batch")
# Without any job profile specified, the default JobProfile object will be used.
# The preferred method is to define a Submission type object to pass to submit()
exampleProfile = JobProfile(cpus=2, memory="4G", qos="batch",email.complete="matt.bradley@jax.org")
exampleSub = Submission(script.file="example.R", cluster="login.sumner.jax.org", profile=exampleProfile)
submit(exampleSub, job.name="example")
#This will save a Submission type object named "example" after completing successfully.
# Alternatively, submit.default will create a new Submission object using passed parameters. It is important
# to note that job.name is required to trigger the default method. In this example we will also define
# sbatch arguments using the slurm.args argument. Since save.sub=TRUE, this will create a new Submission
# type object in .GlobalEnv named "example".
submit("example", script.file="example.R", cluster="login.sumner.jax.org",
slurm.args="-c 2 --mem=4G, -q batch --mail-type=END --mail-user=matt.bradley@jax.org",
save.sub=TRUE)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.