3 ms·Error Propagation in Scientific Computing5 points by ridersofnorth 13y agom_class 13y agoIs there any generic way to preserve loss of numerical accuracy for any given function?