An example step workflow that does a whole bunch of simple random matrix inversion.
This example step workflow generates random matrices, inverts them, does a cumulative sum over an axis, the finally plots them.
By default the workflow and all sub tasks are configured to run in a single process. However, there are example
steps called mapped_raw
and mapped_invert
to give an idea of how to switch from single threaded / process
to parallel data gathering and processing.
If you want to run this in a distributed fashion be sure install the distributed dependencies
(pip install -e .[distributed]
) and additionally create a workflow_config.json
file with the following contents:
{
"project_local_staging_dir": "/allen/aics/modeling/{your_name}/results/example_step_workflow/"
}
Note: The path your provide in the JSON file can point to wherever you want, the path simply must be available to all the workers.
pip install git+https://github.com/AllenCellModeling/example_step_workflow.git
See CONTRIBUTING.md for information related to developing the code.
pip install -e .[dev]
Free software: Allen Institute Software License