Finish up TODOs in QV types

[stelmo]

No description provided.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (5773631) to head (4920f59).
Report is 5 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff            @@
##            master       #58   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            8         8           
  Lines          428       428           
=========================================
  Hits           428       428           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[stelmo]

/format

[github-actions]

✔️ Auto-formatting triggered by this comment succeeded, commited as 6acc986

[stelmo]

Hmm getting a weird bug:

julia> t = LinearValueT(Expression(2)) + QuadraticValueT(Expression(2))
ERROR: MethodError: no method matching QuadraticValueT{FastDifferentiation.Node}(::LinearValueT{FastDifferentiation.Node})
The type `QuadraticValueT{FastDifferentiation.Node}` exists, but no method is defined for this combination of argument types when trying to construct it.

Closest candidates are:
  (::Type{QuadraticValueT{T}} where T)(::Any, ::Any)
   @ ConstraintTrees C:\Users\stelmo\.julia\packages\ConstraintTrees\AhePm\src\quadratic_value.jl:34
  QuadraticValueT{T}(; idxs, weights) where T
   @ ConstraintTrees C:\Users\stelmo\.julia\packages\ConstraintTrees\AhePm\src\quadratic_value.jl:33

Stacktrace:
 [1] +(a::LinearValueT{FastDifferentiation.Node}, b::QuadraticValueT{FastDifferentiation.Node})
   @ ConstraintTrees C:\Users\stelmo\.julia\packages\ConstraintTrees\AhePm\src\quadratic_value.jl:100        
 [2] top-level scope
   @ REPL[30]:1

[exaexa]

I removed some (many) explicit UnionAlls.

It seems to work now at least for:

julia> C.LinearValueT(1) - C.QuadraticValueT(2)
ConstraintTrees.QuadraticValueT{Int64}([(0, 0)], [-1])

If something still fails we should have a test for it (so that there's something to make sure it stays fixed).

[exaexa]

PS we could do type promotion at the place for + but let's avoid that for now unless we find an actual reason.

[exaexa]

LGTM but I expect this won't be sufficient.

[stelmo]

Merging, and I will report back as I find stuff in Diffmet