Get platform-appropriate tolerance for numerical comparisons Addresses Issue #297: Windows CI precision differences
On Windows, floating-point precision can vary due to: - Different compiler optimization flags - Different math libraries (MSYS2 vs native) - Different CPU instruction sets
This function provides a systematic way to handle platform precision differences while maintaining test accuracy.
Type | Intent | Optional | Attributes | Name | ||
---|---|---|---|---|---|---|
real(kind=real64), | intent(in) | :: | base_tolerance |