3 ms·
I actually do this quite regularly in Julia: it provides a BigFloat type which wraps the MPFR library. Since it has generic methods, you can simply do something
by simonbyrne 12y ago
I actually do this quite regularly in Julia: it provides a BigFloat type which wraps the MPFR library. Since it has generic methods, you can simply do something like foo(big(x)), and all the internal computations will be done in extended precision, and the result can be compared to foo(x).
- amelius 12y agoDoes this also work with matrix type of operations? For example solving a linear system?
- simonbyrne 12y agoYes, you can define an Array of BigFloats, and most of the linear algebra operations have general purpose fallbacks that work as expected.