Do I lose performance by (ab)using multivariate polynomials over only one variable?

[Mao Zeng]

I'm writing some code which uses Nemo for polynomial arithmetic with any number of variables. For convenience, I haven't distinguished the multivariate and univariate case. So when there is only one variable, the code is essentially

`R, (x,) = PolynomialRing(QQ, ["x"])`

Do I lose performance by not using

`R, x = PolynomialRing(QQ, "x")`

instead?

It seems that the former code uses `fmpq_mpoly` from FLINT, which may not be optimal for the univariate case?

[tho...@gmail.com]

Yes, the difference will be noticeable. Depending on how you structured your code, it might be possible to just use the univariate instead?

[Mao Zeng]

OK, I'll just to treat the univariate case as a special case in my code. Thanks for confirming that this is worth doing.