A comparative study of automatically generated and large language model-generated unit tests for piecewise function approximation algorithms