Skip to main content

🏁 How To Run Tuva

The slides below describe our standard approach to setting up multiple data sources on Tuva.

The video below includes a more thorough overview of how to setup your claims data using Tuva.

1. Pre-requisites​

You need to start with claims data in a data warehouse and have dbt installed. dbt is easy to install using any package manager like pip or homebrew. You also need to connect dbt to your data warehouse. You do this by configuring your profile.yml file. dbt has instructions for how to do this which you can find on their docs site.

Ensure you're working with a data warehouse that Tuva supports. We officially and unofficially support several data warehouses and you can find the latest up to date info here

2. dbt Setup​

The very first step is to setup a new dbt project. Once dbt is installed you can do this by executing dbt init <project_name> where you replace <project_name> with the name of your project.

Next you need to configure your dbt_project.yml file in the newly created dbt project. This includes 3 steps:

  • Setting your profile
  • Setting Tuva-specific variables
  • Setting the database and schema where dbt should write data to

See the video above for details on how to do this.

In your dbt_project.yml, if you have only claims data you need to set claims_enabled = true and if you have clinical data you need to set clinical_enabled = true. By default these variables are set to false.

Next you'll need to add a "generate schema" macro to your macros folder in the dbt project. This is optional, but if you don't do this your schema names will all be prefixed with your default schema e.g. "public_" which is typically annoying. dbt has documentation on how to do this or see our video above.

Finally you'll need to import Tuva by creating a packages.yml file and adding the following code:

packages:
 - package: tuva-health/the_tuva_project
   version: [">=0.7.0","<0.8.0"]
 - package: dbt-labs/dbt_utils
   version: [ ">=0.9.2" ]

Then execute dbt deps from the command line to import Tuva.

3. Map Your Data​

The next step is mapping your data to the Tuva Input Layer. Every claims and clinical dataset comes in its own schema (i.e. set of tables and columns). Before you can use Tuva you need to convert your schema to the Tuva Input Layer. Do this by creating models (i.e. SQL files) in your dbt project to transform your data into the Input Layer format. The Mapping Guide provides rules of thumb for how to do this.

4. dbt Build​

Next, run dbt build from the command line to build the entire project. This will create hundreds of data tables in your data warehouse. Your source data will be transformed into the Core Data Model, all Data Marts will be built, and all Terminology and Value Sets will be loaded into your data warehouse. This is pretty cool to see with a single command!

5. Data Quality Audit​

Next you need to perform what we call the Data Quality Audit. The DQA is our systematic approach for validating that we've mapped the source data correctly, identifying atomic-level data quality problems, and understanding the impact of those problems on analytics.

6. Explore Data and Docs​

At this point you have now transformed you data into the Tuva data model and are ready to do data analysis.

dbt also comes with awesome documentation built-in. You can run dbt docs generate to generate the docs and then dbt docs serve to serve up the docs to your localhost. A web browser will open when you do this and you can explore the docs.