adjoint(a) != conj(a) for a::Sym

[olof3]

For example,

@syms a
ishermitian([1 a; conj(a) 1])

returns false.

Since Sym <: Number and

@syms b real=true
adjoint(b) == conj(b)

I would have expected this to hold. Seems like the root cause is in SymPy itself :/, since there a.conjugate() != a.adjoint()?

[jverzani]

There was a recent PR that changed @syms and assumptions. Your syntax continues to work with @vars, but for @syms you now have:

julia> @syms a
(a,)

julia> ishermitian([1 a; conj(a) 1])
false

julia> ishermitian([1 a; adjoint(a) 1])
true

julia> @syms a::real
(a,)

julia> ishermitian([1 a; conj(a) 1])
true

Now, following the motto of "no good github issue goes without some improvement" (I'm sure some such proverb exists) I was going to say you should look at the docstring for @syms for more detail, but there is an extra new line that stops the display, so look here (https://github.com/JuliaPy/SymPy.jl/blob/master/src/constructors.jl#L59).

[olof3]

Great, Thank you!