Module tutorial4

source code

tutorial4

Contains all parameters for the evolutionary run, grammar rules, constraints, and specifics about the terminal and function set of the trees in tutorial4. This example file gather all the settings for a 'multiple' polynomial regression. We want to evolve a system of ordered polynomial equations by using the mathematical operators: '+','*' and the ADF: ADF1 and ADF2, in the said order. an ADF2 branch could then call for an ADF1 terminal, but not the opposite. We try to find the following model being shaped as a system of polynomials:

|ADF1=x+y

|ADF2=ADF1^3

from 1 set of 1000 testing data (1000 different values for x and y). This time, each tree does not have to be evaluated 1000 times. We will input a list of 1000 data points in the variable leafs of the tree. There is not much time difference to process 1000 data-points than 5 when we use this trick! Considering the constraints for building the trees, the root node will have 2 children ADF1 and ADF2 (in order), and ADF2 must be able to reuse ADF1 as a terminal... A typical way to run the tutorial would be to:

• modify the settings file so that:

  >>>     # setup for running tutorial 4
          functions = tutorial4.functions
          crossover_mapping=tutorial4.crossover_mapping
          nb_eval=tutorial4.nb_eval
          nb_ex=tutorial4.nb_ex
          ideal_results=tutorial4.GetIdealResultsData()
          terminals =tutorial4.terminals
          Strongly_Typed_Crossover_degree=tutorial4.Strongly_Typed_Crossover_degree
          Substitute_Mutation=tutorial4.Substitute_Mutation
          treeRules = tutorial4.treeRules
          adfOrdered = tutorial4.adfOrdered
          FitnessFunction = tutorial4.FitnessFunction

• use a different fitness function for the evaluation of a list of values instead of a value. We create 2 new fitness functions:
  • EvalTreeForOneListInputSet is a modification of the original EvalTreeForOneInputSet.
  • FinalFitness2 is a modification of FinalFitness that take as input the result returned by the new EvalTreeForOneListInputSet

    These fitness functions are called in FitnessFunction in the tutorial4 module.

    >>>    def FitnessFunction(my_tree):
    >>>        return evalfitness.FinalFitness2(evalfitness.EvalTreeForOneListInputSet(my_tree))

• run the evolution in the main method, and make sure that the root node parameter only have two children.e.g.

  >>>    import evolver
  if __name__ == "__main__":
  >>>    dbname=r'C:\pop_db'
  >>>    evolver.EvolutionRun(2000,(0,2,'root'),2,8,'AddHalfNode',100, 0.00001 ,0.5,0.49,7,0.8,dbname,True)

  This means that we define:

• a population size of 2000 individuals,
• a root node with 2 children
• a minimum tree depth of 2
• a maximum tree depth of 8
• the use of a Ramped half and Half Koza tree building method
• a maximum number of runs of 100 generations before we stop
• a stoping fitness criteria of 0.1 (if the fitness<=0.1, solution found)
• a crossover probability of 0.5
• a mutation probability of 0.49
• a reproduction probability of 0.01 (automatically deduced from the 2 previous values)
• the size of the tournament selection
• the probability of selecting the fitttest during tournament selection
• a database of name and path dbname
• a run done in verbose mode (printing the fittest element of each generation)

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.