These samples scripts show how a subset of task types can be used.
Call the run<TaskType>Task.m scripts to test a given task type.
More generally, the following task types are supported:
See also: Documentation/html/JobWorkflow.html
Demonstrates the use a MATLAB® desktop Docker® image based cluster to run uncompiled
MATLAB code in batch mode as a task. Use an existing running cluster for faster
responses when testing. The cluster will require access to a MATLAB license manager.
Step through runMATLABBatchTask.m to see how the task is created and run.
Shows the use of a task designed to run standalone binary requiring the MATLAB runtime.
Requires compilation on Linux®. Step through runMRTTask.m to see how the task is
created and run. Use an existing running cluster for faster responses when testing.
Imports and executes a Python® notebook from a workspace as a notebook rather than a
script. Step through runNotebookTask.m to see how the task is created and run.
Imports and executes a Python script from a workspace as a script rather than a
notebook. Step through runSparkPythonTask.m to see how the task is created and
run.
Example to run a script on an existing cluster based on its Id.
runSparkPythonTask(cluster='0417-080746-8oy4uhkz')Task creates a number of dependent tasks that form a task graph. Step through runTasksTask.m
to see how multiple tasks are created, linked with dependencies, and executed as
part of a job workflow.