Feature change observer pattern

[Wikunia]

Closes #269

Function notify_constraints_var_changed! calls changed_var! for the constraints that are affected. Currently only the linear constraint implements the function to compute new mins and maxs values.

[codecov[bot]]

Codecov Report

Merging #271 (a1057f1) into master (68fd0b1) will decrease coverage by 0.01%. The diff coverage is 97.87%.

@@            Coverage Diff             @@
##           master     #271      +/-   ##
==========================================
- Coverage   97.62%   97.61%   -0.02%     
==========================================
  Files          52       52              
  Lines        4344     4450     +106     
==========================================
+ Hits         4241     4344     +103     
- Misses        103      106       +3     

Impacted Files	Coverage Δ
src/pruning.jl	100.00% <ø> (ø)
src/types.jl	99.00% <ø> (ø)
src/constraints/activator_constraints.jl	89.36% <55.55%> (-10.64%)	:arrow_down:
src/ConstraintSolver.jl	98.35% <94.11%> (-0.24%)	:arrow_down:
src/Variable.jl	100.00% <100.00%> (ø)
src/constraints.jl	100.00% <100.00%> (ø)
src/constraints/all_different.jl	100.00% <100.00%> (ø)
src/constraints/all_equal.jl	94.94% <100.00%> (ø)
src/constraints/and.jl	100.00% <100.00%> (ø)
src/constraints/boolset.jl	100.00% <100.00%> (ø)
... and 19 more

---

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 68fd0b1...a1057f1. Read the comment docs.

[Wikunia]

Benchmark Report for ConstraintSolver

Job Properties

• Time of benchmarks:
  • Target: 11 Sep 2021 - 13:29
  • Baseline: 11 Sep 2021 - 13:43
• Package commits:
  • Target: 82b89f
  • Baseline: 2d857b
• Julia commits:
  • Target: 6aaede
  • Baseline: 6aaede
• Julia command flags:
  • Target: -O3
  • Baseline: -O3
• Environment variables:
  • Target: None
  • Baseline: None

Results

A ratio greater than 1.0 denotes a possible regression (marked with :x:), while a ratio less than 1.0 denotes a possible improvement (marked with :white_check_mark:). Only significant results - results that indicate possible regressions or improvements - are shown below (thus, an empty table means that all benchmark results remained invariant between builds).

ID	time ratio	memory ratio
["eternity", "5x5_all"]	1.12 (5%) :x:	1.06 (1%) :x:
["eternity", "5x5_opt"]	1.09 (5%) :x:	1.03 (1%) :x:
["eternity", "5x5_opt_ind"]	1.12 (5%) :x:	1.04 (1%) :x:
["eternity", "5x5_opt_rei"]	1.08 (5%) :x:	1.04 (1%) :x:
["eternity", "6x5"]	1.06 (5%) :x:	1.06 (1%) :x:
["eternity", "6x5_ABS"]	1.05 (5%) :x:	1.06 (1%) :x:
["graph_coloring", "US_50colors"]	1.27 (5%) :x:	1.15 (1%) :x:
["graph_coloring", "US_50colors+equal"]	1.23 (5%) :x:	1.16 (1%) :x:
["graph_coloring", "US_50colors+equal_ABS"]	1.35 (5%) :x:	1.18 (1%) :x:
["graph_coloring", "US_8+equal"]	1.09 (5%) :x:	1.06 (1%) :x:
["graph_coloring", "US_8+equal_ABS"]	1.12 (5%) :x:	1.05 (1%) :x:
["graph_coloring", "le450_5d"]	1.31 (5%) :x:	1.36 (1%) :x:
["graph_coloring", "queen7_7"]	1.18 (5%) :x:	1.14 (1%) :x:
["killer_sudoku", "niall_5500"]	0.48 (5%) :white_check_mark:	0.54 (1%) :white_check_mark:
["killer_sudoku", "niall_5500_ABS"]	12.26 (5%) :x:	12.45 (1%) :x:
["killer_sudoku", "niall_5500_special"]	0.67 (5%) :white_check_mark:	0.72 (1%) :white_check_mark:
["killer_sudoku", "niall_5501"]	0.66 (5%) :white_check_mark:	0.69 (1%) :white_check_mark:
["killer_sudoku", "niall_5501_ABS"]	0.05 (5%) :white_check_mark:	0.06 (1%) :white_check_mark:
["killer_sudoku", "niall_5501_special"]	2.12 (5%) :x:	2.21 (1%) :x:
["lp", "issue_83"]	0.94 (5%) :white_check_mark:	1.01 (1%)
["scheduling", "all_different_except_0_len10"]	1.42 (5%) :x:	1.32 (1%) :x:
["scheduling", "all_different_except_0_len12"]	1.36 (5%) :x:	1.31 (1%) :x:
["scheduling", "furniture_moving"]	2.74 (5%) :x:	2.43 (1%) :x:
["scheduling", "organize_day"]	1.43 (5%) :x:	1.62 (1%) :x:
["steiner", "steiner_7"]	493.77 (5%) :x:	304.29 (1%) :x:
["sudoku", "25x25_89"]	1.01 (5%)	1.02 (1%) :x:
["sudoku", "25x25_90"]	1.00 (5%)	1.02 (1%) :x:
["sudoku", "top95_1"]	1.05 (5%)	1.06 (1%) :x:
["sudoku", "top95_11"]	0.99 (5%)	1.04 (1%) :x:
["sudoku", "top95_16"]	1.02 (5%)	1.04 (1%) :x:
["sudoku", "top95_21"]	1.01 (5%)	1.06 (1%) :x:
["sudoku", "top95_26"]	0.93 (5%) :white_check_mark:	1.06 (1%) :x:
["sudoku", "top95_31"]	1.01 (5%)	1.05 (1%) :x:
["sudoku", "top95_36"]	1.04 (5%)	1.06 (1%) :x:
["sudoku", "top95_41"]	1.04 (5%)	1.05 (1%) :x:
["sudoku", "top95_46"]	1.03 (5%)	1.05 (1%) :x:
["sudoku", "top95_51"]	0.95 (5%)	1.05 (1%) :x:
["sudoku", "top95_56"]	1.03 (5%)	1.04 (1%) :x:
["sudoku", "top95_6"]	1.02 (5%)	1.06 (1%) :x:
["sudoku", "top95_61"]	0.96 (5%)	1.05 (1%) :x:
["sudoku", "top95_66"]	0.98 (5%)	1.04 (1%) :x:
["sudoku", "top95_71"]	1.04 (5%)	1.05 (1%) :x:
["sudoku", "top95_76"]	0.98 (5%)	1.05 (1%) :x:
["sudoku", "top95_81"]	0.99 (5%)	1.05 (1%) :x:
["sudoku", "top95_86"]	1.01 (5%)	1.05 (1%) :x:
["sudoku", "top95_91"]	1.00 (5%)	1.05 (1%) :x:

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

• ["eternity"]
• ["graph_coloring"]
• ["killer_sudoku"]
• ["lp"]
• ["scheduling"]
• ["steiner"]
• ["sudoku"]

Julia versioninfo

Target

Julia Version 1.6.1
Commit 6aaedecc44 (2021-04-23 05:59 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  uname: Linux 5.13.12-arch1-1 #1 SMP PREEMPT Wed, 18 Aug 2021 20:49:03 +0000 x86_64 unknown
  CPU: AMD Ryzen 7 PRO 4750U with Radeon Graphics: 
                 speed         user         nice          sys         idle          irq
       #1-16  1700 MHz    1775772 s       6259 s     840826 s    2557498 s          0 s

  Memory: 38.41743850708008 GB (1495.8359375 MB free)
  Uptime: 969387.0 sec
  Load Avg:  4.03  3.62  3.33
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-11.0.1 (ORCJIT, znver2)

Baseline

Julia Version 1.6.1
Commit 6aaedecc44 (2021-04-23 05:59 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  uname: Linux 5.13.12-arch1-1 #1 SMP PREEMPT Wed, 18 Aug 2021 20:49:03 +0000 x86_64 unknown
  CPU: AMD Ryzen 7 PRO 4750U with Radeon Graphics: 
                 speed         user         nice          sys         idle          irq
       #1-16  2029 MHz    1787664 s       6259 s     845095 s    2675998 s          0 s

  Memory: 38.41743850708008 GB (1143.59375 MB free)
  Uptime: 970233.0 sec
  Load Avg:  1.71  2.41  2.88
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-11.0.1 (ORCJIT, znver2)

[Wikunia]

Todo: Compare steiner to before. Probably the change to reverse the complement constraint made this slow. Those reverses should only happen if the constraint was called.

[Wikunia]

Idea: For prune_constraint and still_feasible have a function in between which checks if it's the first node call (which is an attribute of Constraint) if that's the case then it calls a new function first_node_call which defaults to doing nothing. Then sets the attribute to false.

linear_constraints will be the only constraint so far to implement first_node_call which will recompute the extrema

[Wikunia]

Okay we're back to steiner being very slow but hopefully the bug is fixed :smile:

[Wikunia]

Run benchmark tests again to be sure and investigate steiner again...