emilyriederer / dbt-convo-covid

Demo repo with full code described in blog post

dbt demo

This repository demonstrates how to use dbt to publish data that adheres to a controlled vocabulary naming convention.

To motivate the example, it pulls from Google BigQuery public datasets to create a table to evaluate the accuracy of a COVID forecasting model. The author is not an expert on COVID. This repository is very much not intended to make any statements about COVID models or to demonstrate best practices of model monitoring. This is simply an example using a set of publicly available datasets in a database on an understandable and widely relatable topic.

To understand the overall file structure, see the dbt documentation

Useful resources specific to this project are found in:

• macros
  • get_column_names(): Returns all column names for a relation as a list
  • get_matches(): Returns all elements of a list matching a regex
• models
  • model_monitor.sql: Demonstrates the use of the above two macros`
• tests
  • All custom tests are intended to check for controlled vocabulary "contract" adherence