Skip to content

README.md

Latest commit

 

History

History
127 lines (96 loc) · 9.47 KB

README.md

File metadata and controls

127 lines (96 loc) · 9.47 KB

Automated Test Suite Documentation

This project includes a comprehensive suite of tests to validate the functionality, robustness, and reliability of key components. This document provides a categorized overview of the tests, including their purpose and expected outcomes.

Table of Contents

Test Categories and Coverage

1. REST API Tests

Located in: tests/test_rest_api.py

Test Cases:

Test Name Purpose Expected Outcome
test_predict_success Tests the /predict endpoint with valid input data. Returns a 200 OK response with a valid prediction.
test_predict_validation_error Validates that the /predict endpoint handles missing or invalid input fields correctly. Returns a 400 Bad Request response with detailed validation errors.
test_predict_missing_payload Ensures the /predict endpoint handles empty or missing JSON payloads. Returns a 422 Unprocessable Entity response with an appropriate error message.
test_root_endpoint Verifies the root (/) endpoint functionality. Returns a 200 OK response with a health check message.

2. Data Ingestion Tests

Located in: tests/test_services/test_data_ingestion.py

Test Cases:

Test Name Purpose Expected Outcome
test_data_ingestion_with_missing_file Ensures the data ingestion process handles missing input files gracefully. Raises a CustomException caused by a FileNotFoundError, with a message referencing the missing file.
test_data_ingestion_creates_files Validates that data ingestion correctly creates train and test files from the input dataset. Generated train and test files exist in the specified paths and contain valid data.

3. Data Transformation Tests

Located in: tests/test_services/test_data_transformation.py

Test Cases:

Test Name Purpose Expected Outcome
test_data_transformation Ensures the data transformation process works as expected with valid inputs. Returns transformed train and test arrays with expected shapes and saves the preprocessor object.
test_data_transformation_missing_columns Validates that the transformation process handles missing required columns appropriately. Raises a CustomException indicating the missing columns.

4. Model Selection Tests

Located in: tests/test_services/test_model_selection.py

Test Cases:

Test Name Purpose Expected Outcome
test_initiate_model_trainer_success Validates the selection and evaluation of the best model from the given data. Ensures a valid model report is returned, the best model is identified, and its score meets expectations.
test_no_model_meets_threshold Ensures an exception is raised when no model meets the minimum performance threshold. Raises a CustomException with a message indicating no best model was found.
test_invalid_data_handling Tests the behavior of model selection when invalid training and testing data are provided. Returns a model report with None scores for all but Linear Regression, which should score 1.0.

5. Command-Line Argument Tests

Located in: tests/test_command_line.py

Test Cases:

Test Name Purpose Expected Outcome
test_parse_arguments_ingest_monkey Tests argument parsing for the ingest command using monkeypatch. Ensures ingest is recognized and config is properly set.
test_parse_arguments_train_monkey Tests argument parsing for the train command with monkeypatch. Ensures train command is parsed and --debug is correctly set to True.
test_parse_arguments_ingest_mock Tests ingest command parsing using patch. Ensures correct argument assignment for ingest.
test_parse_arguments_train_mock Tests train command parsing using patch. Ensures correct argument assignment for train.
test_parse_arguments_ingest_no_config Ensures ingest works without a specified config file. Confirms default config behavior when none is provided.
test_parse_arguments_train_no_debug Ensures train command behavior when --debug is not passed. Confirms debug defaults to False and correct model config is used.
test_parse_arguments_ingest_with_debug Ensures ingest command accepts the --debug flag. Verifies debug is correctly set to True.
test_parse_arguments_no_command Tests behavior when no command is provided. Ensures the application exits with an error.

Execution Instructions

  1. Ensure all dependencies are installed using:

    pip install -r requirements.txt

  2. Run the tests using pytest:

    PYTHONPATH=$(pwd) pytest -v

Running Specific Tests

To run a specific test module, use:

PYTHONPATH=$(pwd) pytest -v tests/<test_module_name>.py

For example, to run only the command-line argument tests:

PYTHONPATH=$(pwd) pytest -v tests/test_command_line.py

Generating a Test Coverage Report

To generate a test coverage report:

PYTHONPATH=$(pwd) pytest --cov=src --cov-report=html