v0.3.0
SeaPearl v0.3.0
Closed issues:
- DP model vs CP model (#25)
- Change the way evaluation of a learned heursitic is made (#37)
- Add the all different constraint (#49)
- CPModification not correctly updated in presence of views (#55)
- Add a LatinSquares generator (#62)
- Make Lexicographic the default value of a BasicHeuristic (#63)
- Adding common generic value ordering heuristics (#64)
- Unpractical type casting for Vector Constraints (#118)
- Bug in the TSPTW test (#122)
- prunedDomains update check (#128)
- metricsFun can not handle more than 1 classic heuristic (#134)
- Possible TableCT optimisation (#137)
- Current base NN dense layer doesn't use any activation functions.... (#155)
Merged pull requests:
- Feature/alldifferent (#115) (@PierreTsr)
- CompatHelper: bump compat for "Distributions" to "0.25" (#116) (@github-actions[bot])
- Fix for Absolute.jl (#117) (@PierreTsr)
- Table constraint: compact table (#119) (@PierreTsr)
- Fix/vector_casting (#120) (@PierreTsr)
- SeaPearl cleanup and bug fix for RL value selection (#121) (@3rdCore)
- updated README.md (#123) (@3rdCore)
- fix: SetDiffSingleton propagation (#129) (@PierreTsr)
- a prettier SeaPearl (#130) (@PierreTsr)
- iLDS (#133) (@3rdCore)
- all diff for arrays (#135) (@tomsander1998)
- Compact Table optimization (#138) (@PierreTsr)
- Feature/disjunctive (#139) (@kimriouxparadis)
- CPModification upgrade (#140) (@PierreTsr)
- Entire refacto of search metrics (#141) (@3rdCore)
- Problem/nqueens (#142) (@tomsander1998)
- Added new plotting features to compare different metrics ( learned, classic, random ... ) (#143) (@3rdCore)
- CompatHelper: add new compat entry for "RollingFunctions" at version "0.6" (#146) (@github-actions[bot])
- Refacto code (#150) (@3rdCore)
- Feature/reward (#152) (@3rdCore)
- Wo graph clean (#154) (@tomsander1998)
- SeaPearl v0.3.0: RL pipeline refactoring and GPU support (#158) (@PierreTsr)
- Fix/dependencies (#159) (@PierreTsr)
- fix CPModification update with View Variables (#160) (@PierreTsr)
- CompatHelper: bump compat for "ProgressBars" to "1" (#161) (@github-actions[bot])
- Restart Based Search and Policy Gradient Method (#162) (@3rdCore)
- Add globalFeatures and test to FullFeaturedCPNN (#163) (@PierreTsr)
- CompatHelper: add new compat entry for "KernelAbstractions" at version "0.6" (#165) (@github-actions[bot])
- upgraded the functor function (#166) (@PierreTsr)
- removed DataFrames deps (#167) (@PierreTsr)
- CompatHelper: add new compat entry for "NNlib" at version "0.7" (#168) (@github-actions[bot])
- CompatHelper: add new compat entry for "CUDAKernels" at version "0.2" (#169) (@github-actions[bot])
- Failure based search, a first step towards an intelligent variable selection heuristic (#171) (@3rdCore)
- Removing GPU usage from Test (#172) (@PierreTsr)
- CompatHelper: add new compat entry for "FillArrays" at version "0.12" (#173) (@github-actions[bot])
- Removed rolling-mean computation (#175) (@PierreTsr)
- CompatHelper: bump compat for "CUDAKernels" to "0.3" (#176) (@github-actions[bot])
- CompatHelper: bump compat for "KernelAbstractions" to "0.7" (#177) (@github-actions[bot])
- Fix unneeded docstring + inclusion bug (#178) (@PierreTsr)
- Episode redef (#179) (@3rdCore)
- Bug fix on evaluator testset (#182) (@3rdCore)
- Project.toml update (#183) (@PierreTsr)