WIP: Feature MOI

[Wikunia]

Needs at least:

• [x] Test cases
  • [X] Sudoku
  • [X] Killer
  • [x] Graph coloring
• [x] Able to set objective
• [x] Updated docs
• [x] Ability for parameters like: logs and backtrack
• [x] Test for maximizing objective
• [x] Benchmarking

[codecov[bot]]

Codecov Report

:exclamation: No coverage uploaded for pull request base (master@1c7cfb2). Click here to learn what that means. The diff coverage is 95.83%.

@@            Coverage Diff            @@
##             master      #51   +/-   ##
=========================================
  Coverage          ?   98.56%           
=========================================
  Files             ?       19           
  Lines             ?     1393           
  Branches          ?        0           
=========================================
  Hits              ?     1373           
  Misses            ?       20           
  Partials          ?        0

Impacted Files	Coverage Δ
src/MOI_wrapper/MOI_wrapper.jl	94.11% <100%> (ø)
src/equal.jl	100% <100%> (ø)
src/MOI_wrapper/constraints.jl	100% <100%> (ø)
src/eq_sum.jl	100% <100%> (ø)
src/ConstraintSolver.jl	97.98% <100%> (ø)
src/all_different.jl	98.34% <100%> (ø)
src/hashes.jl	100% <100%> (ø)
src/MOI_wrapper/results.jl	81.25% <66.66%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 1c7cfb2...fd0fcc9. Read the comment docs.

[Wikunia]

Last step: Compare benchmark runs:

• [x] Sudoku
• [x] Killer sudoku
• [x] Graph coloring

[Wikunia]

Benchmarks are a bit different because JuMP uses a different variable ordering which changes the hashes of the constraints and therefore the ordering they are called. Graph coloring is worse because the objective changed. Additionally there might a bit overhead on sudoku solving. -> Finding larger test cases. #53