I'm trying to sort the planning entities on decreasing difficulty in the local search phase.
I tried to add the "entitySelector" in the config file like the following but it results in a ConversionException:
<localSearch>
<termination>
<maximumUnimprovedStepCount>500</maximumUnimprovedStepCount>
</termination>
<moveListFactory>
<cacheType>PHASE</cacheType> <!-- STEP, PHASE -->
<selectionOrder>RANDOM</selectionOrder>
<moveListFactoryClass>com.abcdl.be.solver.move.factory.ParentChangeMoveFactory</moveListFactoryClass>
<entitySelector>
<cacheType>PHASE</cacheType>
<selectionOrder>SORTED</selectionOrder>
<sorterManner>DECREASING_DIFFICULTY</sorterManner>
</entitySelector>
</moveListFactory>
<acceptor>
<lateAcceptanceSize>400</lateAcceptanceSize>
<entityTabuSize>5</entityTabuSize>
</acceptor>
<forager>
<pickEarlyType>NEVER</pickEarlyType> <!-- FIRST_BEST_SCORE_IMPROVING -->
<acceptedCountLimit>3</acceptedCountLimit>
</forager>
</localSearch>
The following Comparator class in annotated on the domain model :
public class NodeDifficultyComparator implements Comparator<Node>{
public int compare(Node a, Node b) {
return new CompareToBuilder()
.append(a.getResources(), b.getResources()) // the most difficult nodes are the ones who use the most resources
.append(a.getId(), b.getId())
.toComparison();
}
}
Did I choose a wrong placement for the "entitySelector" tag ? Should I do this in an another way ??
Thanks for your help.
Implementation detail: the element <moveListFactory> is defined by the class MoveListFactoryConfig. If you look at the source of that class, it does not have a field called entitySelector, so you cannot nest an <entitySelector> element in it.
There are 2 ways to solve your problem:
Instead of using <moveListFactory>, use <changeMoveSelector>, which does support an <entitySelector> element. You can then delete your custom ParentChangeMoveFactory class. The <changeMoveSelector> has many advantages over a custom MoveListFactory, such as JIT support and much more.
If you have a good reason to have a custom MoveListFactory (for example due to some complex reasons which move filtering can't cover), then adjust your ParentChangeMoveFactory to sort the entity list before generating the moves. Remember to shallow clone that list first, as you don't want your sorting to affect the entity list references by the workingSolution instance. The point is: if you write a custom MoveListFactory you're in total control on how to generate the moves, but you do need to do everything yourself...
Related
I want to experiment with different local search configurations. They all use the same neighborhood, so I defined the unionMoveSelector in the inheritedSolverBenchmark.
<inheritedSolverBenchmark>
<solver>
...
<localSearch>
<termination>
<unimprovedSecondsSpentLimit>30</unimprovedSecondsSpentLimit>
</termination>
<unionMoveSelector>
<changeMoveSelector />
<swapMoveSelector />
<pillarChangeMoveSelector>
<subPillarType>ALL</subPillarType>
</pillarChangeMoveSelector>
<pillarSwapMoveSelector>
<subPillarType>ALL</subPillarType>
</pillarSwapMoveSelector>
</unionMoveSelector>
</localSearch>
</solver>
</inheritedSolverBenchmark>
Then I created various benchmarks to compare:
<solverBenchmark>
<name>Tabu</name>
<solver>
<localSearch>
<localSearchType>TABU_SEARCH</localSearchType>
</localSearch>
</solver>
</solverBenchmark>
<solverBenchmark>
<name>LateAcceptance</name>
<solver>
<localSearch>
<localSearchType>LATE_ACCEPTANCE</localSearchType>
</localSearch>
</solver>
</solverBenchmark>
<solverBenchmark>
<name>GreatDeluge</name>
<solver>
<localSearch>
<localSearchType>GREAT_DELUGE</localSearchType>
</localSearch>
</solver>
</solverBenchmark>
<solverBenchmark>
<name>Tabu_LA</name>
<solver>
<localSearch>
<acceptor>
<entityTabuSize>7</entityTabuSize>
<lateAcceptanceSize>400</lateAcceptanceSize>
</acceptor>
<forager>
<acceptedCountLimit>1000</acceptedCountLimit>
</forager>
</localSearch>
</solver>
</solverBenchmark>
But now it seems that OptaPlanner executes three phases (construction + 2 x local search). I had expected the localSearch configuration from inheritedSolverBenchmark and plannerBenchmark to be merged. Is that not the case?
No, I am afraid it's not.
When inherited, the SolverConfig's phaseList gets list-merged, not overwritten, nor list-element-merged:
Nobody want's overwritten (as that would just ignore what's in the inherited config), so that's not it.
list-merged: some cases want this. I 've used it a lot to define a CH in the inherited config and the LS variants in the single benchmarks.
list-element-merged: what you're asking. Some cases want this.
We can't detect automatically if it's a case that needs list-merged or list-element-merged, so we had to make a choice and we went with list-merged.
Fix: Either declare it more verbosely or look at the Freemarker template support.
I'm new into OptaPlanner and I'm trying to create an as simple as possible app that assigns few employees to some shifts. The only rule is that one employee can be assigned to one shift per day. I wonder if following solver configuration is not enough:
<solver>
<solutionClass>com.test.shiftplanner.ShiftPlanningSolution</solutionClass>
<entityClass>com.test.shiftplanner.ShiftAssignment</entityClass>
<scoreDirectorFactory>
<scoreDefinitionType>HARD_SOFT</scoreDefinitionType>
<scoreDrl>rules.drl</scoreDrl>
</scoreDirectorFactory>
<!-- Solver termination -->
<termination>
<secondsSpentLimit>60</secondsSpentLimit>
</termination>
<constructionHeuristic>
<constructionHeuristicType>FIRST_FIT</constructionHeuristicType>
</constructionHeuristic>
</solver>
because the collection of ShiftAssignment at ShiftPlanningSolution class remains EMPTY even though the Solver.solve() finishes and getBestSolution() returns something. What's more it seems that my rules at rules.drl are not fired at all. I even added a dummy rule just to see if it is triggered:
rule "test"
when
shiftAssignment : ShiftAssignment()
then
System.out.println(shiftAssignment);
end
and it's not fired at all.
So what are my mistakes here? Thanks in advance!
the rule should be doing something with scoreHolder, see docs chapter 5. But despite that, you should see that rule being fired once for every ShiftAssignement instance in your dataset - check if you have any in there.
I am using Optaplanner to solve an installer booking assignment problem, which is a chaining problem, similar to vehicle routing. An installer(vehicle) may have multiple bookings(customer) assigned to it. I need to implement the chain because I need to evaluate one booking in relation to another booking an a particular order.
So, I declared Booking and Installer as #PlanningEntityCollectionProperty in my Solution. Both Booking and Installer implements Standstill. But only in Booking I declared #PlanningVariable for method getPreviousStandstill().
My config:
<solver>
<!--<environmentMode>FAST_ASSERT</environmentMode>-->
<!-- Domain model configuration -->
<solutionClass>InstallationSolution</solutionClass>
<entityClass>Standstill</entityClass>
<entityClass>Booking</entityClass>
<!-- Score configuration -->
<scoreDirectorFactory>
<scoreDefinitionType>HARD_SOFT</scoreDefinitionType>
<!--<easyScoreCalculatorClass>org.optaplanner.examples.cloudbalancing.solver.score.CloudBalancingEasyScoreCalculator</easyScoreCalculatorClass>-->
<!--<easyScoreCalculatorClass>org.optaplanner.examples.cloudbalancing.solver.score.CloudBalancingMapBasedEasyScoreCalculator</easyScoreCalculatorClass>-->
<!--<incrementalScoreCalculatorClass>org.optaplanner.examples.cloudbalancing.solver.score.CloudBalancingIncrementalScoreCalculator</incrementalScoreCalculatorClass>-->
<!--<scoreDrl>com/tmrnd/pejal/opta/solver/fulfillmentScoreRules.drl</scoreDrl>-->
<initializingScoreTrend>ONLY_DOWN</initializingScoreTrend>
<!--<assertionScoreDirectorFactory>-->
<!--<easyScoreCalculatorClass>org.optaplanner.examples.cloudbalancing.solver.score.CloudBalancingMapBasedEasyScoreCalculator</easyScoreCalculatorClass>-->
<!--</assertionScoreDirectorFactory>-->
</scoreDirectorFactory>
<!-- Optimization algorithms configuration -->
<termination>
<!-- <secondsSpentLimit>20</secondsSpentLimit>-->
<unimprovedSecondsSpentLimit>15</unimprovedSecondsSpentLimit>
</termination>
<constructionHeuristic>
<constructionHeuristicType>FIRST_FIT_DECREASING</constructionHeuristicType>
</constructionHeuristic>
<localSearch>
<unionMoveSelector>
<changeMoveSelector/>
<swapMoveSelector/>
<subChainChangeMoveSelector>
<selectReversingMoveToo>true</selectReversingMoveToo>
</subChainChangeMoveSelector>
<subChainSwapMoveSelector>
<selectReversingMoveToo>true</selectReversingMoveToo>
</subChainSwapMoveSelector>
</unionMoveSelector>
<acceptor>
<entityTabuSize>20</entityTabuSize>
</acceptor>
<forager>
<acceptedCountLimit>1000</acceptedCountLimit>
</forager>
</localSearch>
</solver>
When I try to solve(), I got one installer with all bookings asssigned to him. What do I need to do to improve the situation?
My initial guess would be you have only 1 anchor (1 installer) known to OptaPlanner. Check if the Solution's List's size is bigger than 1 in size. Then check if getter has #ValueRangeProvider and if the id of that ValueRangeProvider is added in the #PlanningEntity's valueRangeProviderRefs.
Also be aware of a common pitfall that might apply here too (but shouldn't if you use value range providers properly): list.add(list) instead of list.addAll(list).
I would love to add support for Working Sets for my Eclipse plugin that used the Common Navigator framework.
In Eclipse bugzilla there is mention that this is supported
None of the online manuals for the Common Navigator explain how to do it
I do not know where to start even since there is no extension point for it, and the Working Set implementation classes are all "internal". I have a very basic navigator setup showing default project resources and some additional IFileSystem stuff implementing ITreeContentProvider.
You can get the working set manager using:
IWorkingSetManager manager = PlatformUI.getWorkbench().getWorkingSetManager();
and from that get the visible working sets with:
IWorkingSet [] workingSets = manager.getWorkingSets();
the members of a working set can be accessed with:
IAdaptable [] elements = workingSet.getElements();
so you could use the working sets list as the input for the tree viewer and adjust your tree content provider to deal with this.
In retrospect the following is a better solution. Instead of implementing ITreeContentProvider and traversing the working sets ourselves, we can reuse existing standard providers for the same content, which might work better.
You can use them like so:
<extension
point="org.eclipse.ui.navigator.viewer">
<viewerContentBinding
viewerId="rascal.navigator">
<includes>
<contentExtension pattern="org.eclipse.ui.navigator.resourceContent" />
<contentExtension pattern="org.eclipse.ui.navigator.resources.filters.*"/>
<contentExtension pattern="org.eclipse.ui.navigator.resources.linkHelper"/>
<contentExtension pattern="org.eclipse.ui.navigator.resources.workingSets"/>
</includes>
</viewerContentBinding>
In particular the org.eclipse.ui.navigator.resources.workingSets is what adds working sets capabilities to your navigator.
Adding your own content then becomes an issue of adding another content provider which ignores workingsets and projects and other kinds of resources which are already taken care of, e.g. like so:
<extension
point="org.eclipse.ui.navigator.navigatorContent">
<navigatorContent
activeByDefault="true"
contentProvider="org.rascalmpl.eclipse.navigator.NavigatorContentProvider"
id="org.rascalmpl.navigator.searchPathContent"
labelProvider="org.rascalmpl.eclipse.navigator.NavigatorContentLabelProvider"
name="Rascal search path"
priority="normal">
<triggerPoints>
<or>
<instanceof value="org.eclipse.core.resources.IResource"/>
</or>
</triggerPoints>
<possibleChildren>
<or>
<instanceof value="java.lang.Object"/>
</or>
</possibleChildren>
<actionProvider
class="org.rascalmpl.eclipse.navigator.NavigatorActionProvider"
id="org.rascalmpl.navigator.actions">
</actionProvider>
<commonSorter
class="org.rascalmpl.eclipse.navigator.Sorter">
</commonSorter>
</navigatorContent>
<commonWizard
type="new"
wizardId="rascal_eclipse.wizards.NewRascalFile">
<enablement></enablement>
</commonWizard>
<commonWizard
type="new"
wizardId="rascal_eclipse.projectwizard">
<enablement></enablement>
</commonWizard>
</extension>
and in the NavigatorContentProvider class we implement getElements and getChildren but only for our own additional content.
For simplicity, say I have a resource users. The HTTP call GET users/ returns a list of links to concrete users:
<users>
<link rel='user' href='/users/user/1/'/>
<link rel='user' href='/users/user/2/'/>
<link rel='user' href='/users/user/3/'/>
....
</users>
The result representation is described in a specific media type:
application/vnd.company.Users+xml
In our frontends, we want to display a table with all users. This means we need to be able to fetch user information to display, such as the name, gender, friends, ... I would like to avoid that we need a separate request for each user (GET /users/user/x/) to retrieve this information. In addition, some frontends will only display the name, while other frontends will display the name and his/her friends. And so on.
In essence, we are still returning users, but with extentions depending on what the frontend needs.
Which option would you choose? Why?
(1) Make GET users/ customizable via parameters such that the customizations are listed. Depending on the customizations , different media types might be returned, since the syntax of one version/combination might be very different than one of another version/combination:
GET users/ -> application/vnd.company.Users+xml
GET users/?fields=name,gender -> application/vnd.company.Users+xml
GET users/?fields=name,gender,friends -> application/vnd.company.UsersWithFriends+xml
(2) Different resources are created to distinguish different between media types. Parameters are still used for basic customizations covered by the media type. This gives:
GET users?fields=name -> application/vnd.company.Users+xml
GET users?fields=name,gender -> application/vnd.company.Users+xml
GET users_with_friends?fields=gender -> application/vnd.company.UsersWithFriends+xml
(3) The same as (1), but instead of parameters, the desired media type is set by the client in the Accept header. Customizable fields covered by the media type are still set via parameters:
GET users/?fields=name ACCEPT application/vnd.company.Users+xml
GET users/?fields=name,gender ACCEPT application/vnd.company.Users+xml
GET users/?fields=name,gender ACCEPT application/vnd.company.UsersWithFriends+xml
(4) Something else?
To answer my own question, I think that:
Solution (1) is very very wrong. The media type must not be dependant on parameters.
Solution (2) and (3) are more or less equal and up to preferences. I prefer (3) since this would not introduce an explosion of resources to be introduced. In addition, in essence we are still returning users. The only difference is the amount of information, reflected by different media types, that is returned. So one might argue that there is no real need to introduce new resources as done in (2).
Do you agree? What do you think?
(3) is surely the best using strict Media Type, but would require specific HTTP Request client and won't be accessible through basic URL open library or browser.
Why not using solution 1 with another extra parameter : names "expect" or "as".
ie:
users/?fields=name,gender&expect=application/vnd.company.Users+xml
users/?fields=name,gender&expect=application/vnd.company.UsersWithFriends+xml
This would be the same as ACCEPT solution but won't need very custom client library to forge the request.
However you'll have to parse the parameter to provide correct output (the (3) would also have this requirement for parsing the ACCEPT)
Personally, I'm not a fan of using query string parameters to allow clients to pick the data elements they wish to include in a representation. I find it makes it hard to optimize the server and it pollutes the cache with many overlapping variants. Also, you really shouldn't try and use conneg to select between representations that contain different sets of data. Conneg is really just for selecting the serialization format.
With the Hal media type you can approach this problem a bit differently. Consider a service with a root representation that looks like:
<resource rel="self"
href="http://example.org/userservice"
xmlns:us="http://example.org/userservice/rels">
<link rel="us:users" name="users" href="http://example.org/users">
<link rel="us:userswithfriends" href="http://example.org/userswithfriends">
</resource>
When you use hal, instead of using the media type documentation to describe your application domain, you can use link relations. In this case, the us:users link points to a document that contains a list of users. I know the namespace stuff looks a bit wierd, but it is not really being used as an XML namespace, just as way of making a Compact URI (CURIE). When you invent your own rel values, they need to be specified in the form of a URI to try and ensure uniqueness.
The list of users would look something like:
<resource rel="self"
href="http://example.org/users"
xmlns:us="http://example.org/userservice/rels">
<resource rel="us:user" name="1" href="/user/1">
<name>Bob</name>
<age>45</age>
<resource>
<resource rel="us:user" name="2" href="/user/2">
<name>Fred</name>
<age>Bill</age>
<resource>
</resource>
and 'us:userswithfriends' points to a different resource that contains the list of users with each user containing a list of friends.
<resource rel="self"
href="http://example.org/users"
xmlns:us="http://example.org/userservice/rels">
<resource rel="us:user" name="1" href="/user/1">
<name>Bob</name>
<resource rel="us:friend" name="1" href="/user/10">
<name>Sheila</name>
<resource>
<resource rel="us:friend" name="2" href="/user/74">
<name>Robert</name>
<resource>
<resource>
<resource rel="user" name="2" href="/user/2">
<name>Fred</name>
<resource rel="us:friend" name="1" href="/user/14">
<name>Bill</name>
<resource>
<resource rel="us:friend" name="2" href="/user/33">
<name>Margaret</name>
<resource>
<resource>
</resource>
With hal it is the documentation of your rels (us:users, us:friend) that decribes what data elements are allowed to exist in the resource element. You are free to embed all of the data of the resource, or more likely just a subset of the data. If the client wants to access a completely representation of the embedded resource then it can follow the provided link.