You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Based on interactions with a user (Gunter Malle) today, here are some ideas for making life more comfortable for users. Consider this starting point:
julia> QQAb, E = abelian_closure(QQ)
(Abelian closure of Q, Generator of abelian closure of Q)
julia> x = E(5) + E(5)^4
-ζ(5)^3 - ζ(5)^2 - 1
julia> conj(x)
-ζ(5)^3 - ζ(5)^2 - 1
julia> conj(x) == x
true
implement is_real(:: QQAbFieldElem)
julia> is_real(x)
ERROR: MethodError: no method matching isreal(::QQAbFieldElem{AbsSimpleNumFieldElem})
implement > comparison -- by fixing an embedding into QQBar esp. $\mathbb{C}$, based on discussion with @fieker we could use the "standard embedding" (of course experts will immediately argue there are many embeddings, and of course complex_embeddings and real_embeddings for number fields are a thing and important -- but experts probably won't be using abelian_closure to start with)
implement coercion to QQBar -- this again requires choosing an embedding, but if we do that anyway, we may as well also support it here
make it easy to get a numerical approximation, e.g. by adding coercion to ComplexField / RealField
The text was updated successfully, but these errors were encountered:
Based on interactions with a user (Gunter Malle) today, here are some ideas for making life more comfortable for users. Consider this starting point:
is_real(:: QQAbFieldElem)
implement$\mathbb{C}$ , based on discussion with @fieker we could use the "standard embedding" (of course experts will immediately argue there are many embeddings, and of course
>
comparison -- by fixing an embedding intoQQBar
esp.complex_embeddings
andreal_embeddings
for number fields are a thing and important -- but experts probably won't be usingabelian_closure
to start with)implement coercion to
QQBar
-- this again requires choosing an embedding, but if we do that anyway, we may as well also support it heremake it easy to get a numerical approximation, e.g. by adding coercion to
ComplexField
/RealField
The text was updated successfully, but these errors were encountered: