Components of a Pipeline

A standard pipeline consists of a dataset followed by a series of capsules that write results to a Results Bucket.

Each capsule is a standalone and fully reproducible process that reads data from the data folder and writes results to the results folder. When implemented in a pipeline, the contents of each capsule’s data folder are ignored. The Dataset is passed into the first capsule’s data folder, and the results folder of the first capsule becomes the data folder of the second capsule. Results from each capsule will only be saved if it is connected to the Results Bucket.

This section covers the three main components of a pipeline:

• Capsules

• Datasets

• Results Bucket

In addition to the above, the following will be covered:

• How to customize the pipeline using the Map Paths menu and Capsule Settings