Pipeline Configuration

The pipeline config file pipeline.yaml describes the configuration of your pipeline:

  1. Triggers - which file input file patterns should trigger this pipeline

  2. Ingest Class - class name of the ingest pipeline to use

  3. Dependent Config Files - which yaml files to use for the retriever, dataset, quality management, and storage

Each pipeline template will include a starter pipeline config file in the config folder. It will work out of the box, but the configuration should be tweaked according to the specifics of your pipeline. Consult the Getting Started section for more information on getting started with a template.

Note

To prevent redundancy, Tsdat config files are designed to be shared across multiple pipelines. In the pipeline config file, you can specify a shared config file to use (ie., shared/config/dataset.yaml) and then override specific values in the overrides section.

An annotated example of an ingest pipeline config file is provided below:

 1#  Name of the Ingest Pipeline to use
 2classname: tsdat.pipeline.ingest.IngestPipeline
 3
 4#  Regex patterns that should trigger this pipeline
 5triggers:
 6  - .*example_pipeline.*\.csv
 7
 8# Retriever config
 9retriever:
10  path: pipelines/example_pipeline/config/retriever.yaml
11
12# Dataset config.  In this example, we use a dataset.yaml file that is shared across multiple pipelines,
13# but we override one global attribute specifying a different location and we add one additional variable attribute.
14dataset:
15  path: shared/config/dataset.yaml
16  overrides:
17    /attrs/location_id: sgp
18    /data_vars/first/attrs/new_attribute: please add this attribute
19
20# Quality config - shared across multiple pipelines
21quality:
22  path: shared/config/default-quality.yaml
23
24# Storage config - shared across multiple pipelines
25storage:
26  path: shared/config/storage.yaml