ec.gp.koza
Class CrossoverPipeline

java.lang.Object
  extended by ec.BreedingSource
      extended by ec.BreedingPipeline
          extended by ec.gp.GPBreedingPipeline
              extended by ec.gp.koza.CrossoverPipeline
All Implemented Interfaces:
Prototype, Setup, SteadyStateBSourceForm, RandomChoiceChooser, java.io.Serializable, java.lang.Cloneable

public class CrossoverPipeline
extends GPBreedingPipeline

CrossoverPipeline is a GPBreedingPipeline which performs a strongly-typed version of Koza-style "Subtree Crossover". Two individuals are selected, then a single tree is chosen in each such that the two trees have the same GPTreeConstraints. Then a random node is chosen in each tree such that the two nodes have the same return type. If by swapping subtrees at these nodes the two trees will not violate maximum depth constraints, then the trees perform the swap, otherwise, they repeat the hunt for random nodes.

The pipeline tries at most tries times to a pair of random nodes BOTH with valid swap constraints. If it cannot find any such pairs after tries times, it uses the pair of its last attempt. If either of the nodes in the pair is valid, that node gets substituted with the other node. Otherwise an individual invalid node isn't changed at all (it's "reproduced").

Compatibility with constraints. Since Koza-I/II only tries 1 time, and then follows this policy, this is compatible with Koza. lil-gp either tries 1 time, or tries forever. Either way, this is compatible with lil-gp. My hacked lil-gp kernel either tries 1 time, n times, or forever. This is compatible as well.

This pipeline typically produces up to 2 new individuals (the two newly- swapped individuals) per produce(...) call. If the system only needs a single individual, the pipeline will throw one of the new individuals away. The user can also have the pipeline always throw away the second new individual instead of adding it to the population. In this case, the pipeline will only typically produce 1 new individual per produce(...) call.

Typical Number of Individuals Produced Per produce(...) call
2 * minimum typical number of individuals produced by each source, unless tossSecondParent is set, in which case it's simply the minimum typical number.

Number of Sources
2

Parameters

base.tries
int >= 1
(number of times to try finding valid pairs of nodes)
base.maxdepth
int >= 1
(maximum valid depth of a crossed-over subtree)
base.tree.0
0 < int < (num trees in individuals), if exists
(first tree for the crossover; if parameter doesn't exist, tree is picked at random)
base.tree.1
0 < int < (num trees in individuals), if exists
(second tree for the crossover; if parameter doesn't exist, tree is picked at random. This tree must have the same GPTreeConstraints as tree.0, if tree.0 is defined.)
base.ns.n
classname, inherits and != GPNodeSelector,
or String same
(GPNodeSelector for parent n (n is 0 or 1) If, for ns.1 the value is same, then ns.1 a copy of whatever ns.0 is. Note that the default version has no n)
base.toss
bool = true or false (default)/td>
(after crossing over with the first new individual, should its second sibling individual be thrown away instead of adding it to the population?)

Default Base
gp.koza.xover

Parameter bases

base.ns.n
nodeselectn (n is 0 or 1)

Version:
1.0
Author:
Sean Luke
See Also:
Serialized Form

Field Summary
static int INDS_PRODUCED
           
 int maxDepth
          The deepest tree the pipeline is allowed to form.
 GPNodeSelector nodeselect1
          How the pipeline selects a node from individual 1
 GPNodeSelector nodeselect2
          How the pipeline selects a node from individual 2
static int NUM_SOURCES
           
 int numTries
          How many times the pipeline attempts to pick nodes until it gives up.
static java.lang.String P_CROSSOVER
           
static java.lang.String P_MAXDEPTH
           
static java.lang.String P_NUM_TRIES
           
static java.lang.String P_TOSS
           
 GPIndividual[] parents
          Temporary holding place for parents
 boolean tossSecondParent
          Should the pipeline discard the second parent after crossing over?
 int tree1
          Is the first tree fixed? If not, this is -1
 int tree2
          Is the second tree fixed? If not, this is -1
 
Fields inherited from class ec.gp.GPBreedingPipeline
P_NODESELECTOR, P_TREE, TREE_UNFIXED
 
Fields inherited from class ec.BreedingPipeline
DYNAMIC_SOURCES, mybase, P_NUMSOURCES, P_SOURCE, sources, V_SAME
 
Fields inherited from class ec.BreedingSource
CHECKBOUNDARY, DEFAULT_PRODUCED, NO_PROBABILITY, P_PROB, probability, UNUSED
 
Constructor Summary
CrossoverPipeline()
           
 
Method Summary
 java.lang.Object clone()
          Creates a new individual cloned from a prototype, and suitable to begin use in its own evolutionary context.
 Parameter defaultBase()
          Returns the default base for this prototype.
 int numSources()
          Returns the number of sources to this pipeline.
 int produce(int min, int max, int start, int subpopulation, Individual[] inds, EvolutionState state, int thread)
          Produces n individuals from the given subpopulation and puts them into inds[start...start+n-1], where n = Min(Max(q,min),max), where q is the "typical" number of individuals the BreedingSource produces in one shot, and returns n.
 void setup(EvolutionState state, Parameter base)
          Sets up the BreedingPipeline.
 int typicalIndsProduced()
          Returns 2 * minimum number of typical individuals produced by any sources, else 1* minimum number if tossSecondParent is true.
 boolean verifyPoints(GPInitializer initializer, GPNode inner1, GPNode inner2)
          Returns true if inner1 can feasibly be swapped into inner2's position.
 
Methods inherited from class ec.gp.GPBreedingPipeline
produces
 
Methods inherited from class ec.BreedingPipeline
finishProducing, individualReplaced, maxChildProduction, minChildProduction, preparePipeline, prepareToProduce, sourcesAreProperForm
 
Methods inherited from class ec.BreedingSource
getProbability, pickRandom, setProbability, setupProbabilities
 
Methods inherited from class java.lang.Object
equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

P_NUM_TRIES

public static final java.lang.String P_NUM_TRIES
See Also:
Constant Field Values

P_MAXDEPTH

public static final java.lang.String P_MAXDEPTH
See Also:
Constant Field Values

P_CROSSOVER

public static final java.lang.String P_CROSSOVER
See Also:
Constant Field Values

P_TOSS

public static final java.lang.String P_TOSS
See Also:
Constant Field Values

INDS_PRODUCED

public static final int INDS_PRODUCED
See Also:
Constant Field Values

NUM_SOURCES

public static final int NUM_SOURCES
See Also:
Constant Field Values

nodeselect1

public GPNodeSelector nodeselect1
How the pipeline selects a node from individual 1


nodeselect2

public GPNodeSelector nodeselect2
How the pipeline selects a node from individual 2


tree1

public int tree1
Is the first tree fixed? If not, this is -1


tree2

public int tree2
Is the second tree fixed? If not, this is -1


numTries

public int numTries
How many times the pipeline attempts to pick nodes until it gives up.


maxDepth

public int maxDepth
The deepest tree the pipeline is allowed to form. Single terminal trees are depth 1.


tossSecondParent

public boolean tossSecondParent
Should the pipeline discard the second parent after crossing over?


parents

public GPIndividual[] parents
Temporary holding place for parents

Constructor Detail

CrossoverPipeline

public CrossoverPipeline()
Method Detail

defaultBase

public Parameter defaultBase()
Description copied from interface: Prototype
Returns the default base for this prototype. This should generally be implemented by building off of the static base() method on the DefaultsForm object for the prototype's package. This should be callable during setup(...).


numSources

public int numSources()
Description copied from class: BreedingPipeline
Returns the number of sources to this pipeline. Called during BreedingPipeline's setup. Be sure to return a value > 0, or DYNAMIC_SOURCES which indicates that setup should check the parameter file for the parameter "num-sources" to make its determination.

Specified by:
numSources in class BreedingPipeline

clone

public java.lang.Object clone()
Description copied from interface: Prototype
Creates a new individual cloned from a prototype, and suitable to begin use in its own evolutionary context.

Typically this should be a full "deep" clone. However, you may share certain elements with other objects rather than clone hem, depending on the situation:

Implementations.

Specified by:
clone in interface Prototype
Overrides:
clone in class BreedingPipeline

setup

public void setup(EvolutionState state,
                  Parameter base)
Description copied from class: BreedingSource
Sets up the BreedingPipeline. You can use state.output.error here because the top-level caller promises to call exitIfErrors() after calling setup. Note that probability might get modified again by an external source if it doesn't normalize right.

The most common modification is to normalize it with some other set of probabilities, then set all of them up in increasing summation; this allows the use of the fast static BreedingSource-picking utility method, BreedingSource.pickRandom(...). In order to use this method, for example, if four breeding source probabilities are {0.3, 0.2, 0.1, 0.4}, then they should get normalized and summed by the outside owners as: {0.3, 0.5, 0.6, 1.0}.

Specified by:
setup in interface Prototype
Specified by:
setup in interface Setup
Overrides:
setup in class BreedingPipeline
See Also:
Prototype.setup(EvolutionState,Parameter)

typicalIndsProduced

public int typicalIndsProduced()
Returns 2 * minimum number of typical individuals produced by any sources, else 1* minimum number if tossSecondParent is true.

Overrides:
typicalIndsProduced in class BreedingPipeline

verifyPoints

public final boolean verifyPoints(GPInitializer initializer,
                                  GPNode inner1,
                                  GPNode inner2)
Returns true if inner1 can feasibly be swapped into inner2's position.


produce

public int produce(int min,
                   int max,
                   int start,
                   int subpopulation,
                   Individual[] inds,
                   EvolutionState state,
                   int thread)
Description copied from class: BreedingSource
Produces n individuals from the given subpopulation and puts them into inds[start...start+n-1], where n = Min(Max(q,min),max), where q is the "typical" number of individuals the BreedingSource produces in one shot, and returns n. max must be >= min, and min must be >= 1. For example, crossover might typically produce two individuals, tournament selection might typically produce a single individual, etc.

Specified by:
produce in class BreedingSource