Practice: Grouping dbt models by layer

Override the get_group_name method in your CustomizedDagsterDbtTranslator to group each dbt model by their layer (marts and staging).

Hint: dbt_resource_props is a Python dictionary with a structure that similar to the following:

{
    "database": "data",
    "schema": "main",
    "name": "stg_trips",
    "resource_type": "model",
    "package_name": "analytics",
    "path": "staging/stg_trips.sql",
    "original_file_path": "models/staging/stg_trips.sql",
    "unique_id": "model.analytics.stg_trips",
    "fqn": ["analytics", "staging", "stg_trips"],
    "alias": "stg_trips",
    ... #other properties
}

get_group_name expects to return a string to group the dbt models by. What property of dbt_resource_props can you access (and maybe even index!) to group the models by layer (ex. marts or staging)?

Check your work

The method you built should look similar to the following code. Click View answer to view it.

def get_group_name(self, dbt_resource_props):
    return dbt_resource_props["fqn"][1]