This PR introduces a suite of unit tests for numeric_functions.py located within the within the mathematical functions directory. This PR also resolves issues from the numeric_functions.py file, specifically with functions like Hypotenuse -- hyp(), Negate -- neg(), and Clamp -- clamp().
Bug Resolution: Fixed pointer issues for both hypotenuse and clamp. For hypotenuse, it was missing a parameter feature, and also and out pointer. For clamp, the Boolean for batching was missing for the call_to_clib function. For negate, the only compatible array was of type f32, therefore I added a converter that grabbed the dtype of the inputted array and set that as the standard.
General Functionality: Tests confirm that the library accurately handles tests across various array shapes and dimensions, ensuring these operations perform correctly from scalars up to four-dimensional arrays. This encompasses verifying the maintenance of result dimensions in accordance with input shapes.
Data Type and Edge Case Handling: The library's robustness is tested against a comprehensive set of data types including integers, floating-points, and complex numbers (where applicable), affirming operation compatibility and precision. Additionally, scenarios involving shape mismatches, negative dimensions, and zero-sized arrays are explored, expecting runtime errors to prevent undefined behaviors.
This PR introduces a suite of unit tests for numeric_functions.py located within the within the mathematical functions directory. This PR also resolves issues from the numeric_functions.py file, specifically with functions like Hypotenuse -- hyp(), Negate -- neg(), and Clamp -- clamp().
Bug Resolution: Fixed pointer issues for both hypotenuse and clamp. For hypotenuse, it was missing a parameter feature, and also and out pointer. For clamp, the Boolean for batching was missing for the call_to_clib function. For negate, the only compatible array was of type f32, therefore I added a converter that grabbed the dtype of the inputted array and set that as the standard.
General Functionality: Tests confirm that the library accurately handles tests across various array shapes and dimensions, ensuring these operations perform correctly from scalars up to four-dimensional arrays. This encompasses verifying the maintenance of result dimensions in accordance with input shapes.
Data Type and Edge Case Handling: The library's robustness is tested against a comprehensive set of data types including integers, floating-points, and complex numbers (where applicable), affirming operation compatibility and precision. Additionally, scenarios involving shape mismatches, negative dimensions, and zero-sized arrays are explored, expecting runtime errors to prevent undefined behaviors.