Testing Causality in Scientific Modelling Software

From simulating galaxy formation to viral transmission in a pandemic, scientific models play pivotal role developing theories and supporting government policy decisions that affect us all. Given these critical applications, poor modelling assumption or bug could have far-reaching consequences. However, possess several properties make them notoriously difficult test, including complex input space, long execution times, non-determinism, rendering existing testing techniques impractical. In fields such as epidemiology, where researchers seek answers challenging causal questions, statistical methodology known Causal Inference has addressed similar problems, enabling the inference of conclusions from noisy, biased, sparse data instead costly experiments. This paper introduces Testing Framework: framework uses establish effects data, users conduct software activities concerning effect change, Metamorphic Testing, posteriori . We present three case studies covering real-world models, demonstrating how Framework can infer metamorphic test outcomes reused, confounded provide an efficient solution for software.