dbt_duckdb_python_model_with_pandas_dataframe_filtering.py

1import pandas as pd
2
3def model(dbt, session):
4    # The dbt object provides access to your project's configuration
5    # and upstream models.
6    
7    # Setting the materialization to table (optional, usually set in dbt_project.yml)
8    dbt.config(materialized="table")
9
10    # Use dbt.ref to get a reference to an existing model as a DataFrame
11    # Note: dbt-duckdb automatically passes a DuckDB PyRelation or Pandas DF 
12    # depending on the configuration.
13    upstream_df = dbt.ref("my_upstream_model")
14
15    # Perform standard Python data transformations
16    final_df = upstream_df[upstream_df['column_name'] > 0]
17
18    # The function must return a DataFrame (Pandas, Polars, or PyArrow)
19    return final_df