CorrelationFunction with steady_state = false about quantumcumulants.jl

[ChristophHotter]

@tnadolny thank you for reporting this. @j-moser fixed this in #158. It should work now in version 0.2.15.

[tnadolny]

Thank you!
The new test works well for me. However, I also tried the test without the filter_function, expecting it to work as well.
I get two different errors:
1st error:

u0_c = correlation_u0(corr_sc, sol.u[end])
# Could not find initial value for ⟨σ211*σ222⟩ !

2nd error:

@named csys = ODESystem(corr_sc)
# MethodError: no method matching has_cluster(::Sym{Parameter, Base.ImmutableDict{DataType, Any}})

Regarding the first error, I checked, that ⟨σ211*σ222⟩ is not in the ODESystem:

for s in sys.states
    println(s)
end

var"⟨a′*a⟩"(t)
var"⟨σ221⟩"(t)
var"⟨a′*σ121⟩"(t)
var"⟨a′⟩"(t)
var"⟨a*σ221⟩"(t)
var"⟨σ211*σ122⟩"(t)
var"⟨σ211⟩"(t)
var"⟨a*a⟩"(t)
var"⟨a*σ121⟩"(t)
var"⟨σ221*σ122⟩"(t)
var"⟨σ121*σ122⟩"(t)
var"⟨σ221*σ222⟩"(t)

Is the missing state ⟨σ211*σ222⟩ the same as the conjugate of the state var"⟨σ221*σ122⟩"(t)? If so, could one extend correlation_u0() to check for such permutations?