3 ms·
By using a field extension over √5, you can evaluate this exactly without rounding or resorting to floating point.
by mappu 15y ago
By using a field extension over √5, you can evaluate this exactly without rounding or resorting to floating point.
- euyyn 15y agoI think that's a very sharp observation. I tried it, but in order to exponentiate the vectors (instead of floating point numbers) I end up with the original matrix algorithm (pre-eigenvalues)...