tecton-unit-testing

Rules to follow when you create Tecton FeatureView unit tests

CRITICAL: Before adding a new unit test, make sure that tecton plan works. This ensures the Feature View definitions are at least valid.

Feature retrieval methods such as get_features_in_range and run_transformation use validated Tecton Objects during retrieval by default. Validation requires access to the Tecton API and a Compute Cluster which are often unavailable in Unit Testing environments.

    To use these methods using mock inputs **without validation** in Unit Tests, set the validation mode to `skip` using:
    ```
    conf.set("TECTON_SKIP_OBJECT_VALIDATION", "True")
    ```

In your unit test for a given FeatureView, always fetch a reference to the FeatureView using the TestRepo test fixture. Never try to import it directly from the module. Sample code that shows how to use repo_fixture.get_feature_view to retrieve the unit under test:

### tests/transaction_amount_is_high.py ###
from tecton import TestRepo
import pandas

# Testing the 'transaction_amount_is_high' feature which depends on request data ('amount') as input
def test_transaction_amount_is_high(repo_fixture: TestRepo):
    transaction_amount_is_high = repo_fixture.get_feature_view("transaction_amount_is_high")
    transaction_request = pandas.DataFrame({"amount": [124, 10001, 34235436234]})

    # Use a MockContext to mock any secrets or resources
    mock_context = MockContext(secrets={"my_secret": "my_secret_value"})
    actual = transaction_amount_is_high.run_transformation(
        input_data={
            "transaction_request": transaction_request,
            "context": mock_context,
        },
    ).to_pandas()

    expected = pandas.DataFrame({"transaction_amount_is_high": [0, 1, 1]})
    pandas.testing.assert_frame_equal(actual, expected)

Overview

Tecton supports unit testing for feature definitions to ensure the correctness of feature definitions. Unit tests can be defined in feature repositories in file paths matching the pattern **/tests/*.py.