Pegasus Workflow
We use Pegasus for the workflow on the HCC server. We currently have two workflows.
The first workflow takes the original data and compresses them into zip files to reduce the size and the number of the original data. Here is a workflow diagram illustrating what is above (file/folder names are used as examples):
The second workflow takes a compressed dataset of plant hyperspectral images and extracts a numpy array from it. Then, it creates a predicted image for each using code from https://github.com/freemao/schnablelab. These predict images are run through trait extraction python code that extracts the inflorescence height, inflorescence width, stem height, and plant height from each of the images and outputs them on to a .csv file for that plant. The .csv file is run through a visualization python code that outputs four graphs showing the growth curve for each measurement for each plant onto a .png file.
Here is a workflow diagram illustrating what is above (file/folder names are used as examples):
Users can use the workflow user guide to run the code on the HCC server that is located on the root directory of the repository.