Module es.upm.fi.cig.multictbnc
Package es.upm.fi.cig.multictbnc.writers.performance

Class ExcelExperimentsWriter

java.lang.Object
es.upm.fi.cig.multictbnc.writers.performance.MetricsWriter
es.upm.fi.cig.multictbnc.writers.performance.ExcelExperimentsWriter
public class ExcelExperimentsWriter extends MetricsWriter
Allows writing the results of the experiments in an Excel file.

  • Constructor Details

    • ExcelExperimentsWriter

      public ExcelExperimentsWriter(List<String> scoreFunctions, List<String> nameDatasets, List<String> nameModels, List<String> nameClassVariables, List<String> nameFeatureVariables, BNParameterLearningAlgorithm bnPLA, CTBNParameterLearningAlgorithm ctbnPLA, String penalisationFunction, String initialStructure, List<Long> seeds, String nameGeneratedFile)
      Initialises the writer.
      Parameters:
      scoreFunctions - list of score functions
      nameDatasets - list of dataset names
      nameModels - list of model names
      nameClassVariables - names of class variables
      nameFeatureVariables - names of feature variables
      bnPLA - parameter learning algorithm for a Bayesian network
      ctbnPLA - parameter learning algorithm for a continuous time Bayesian network
      penalisationFunction - name of the penalisation function
      initialStructure - name for the initial structure
      seeds - seeds that are used to shuffle the datasets. It is assumed that each dataset will be evaluated with each of those seeds
      nameGeneratedFile - name of the resulting file

    • ExcelExperimentsWriter

      public ExcelExperimentsWriter(List<String> nameDatasets, List<String> nameModels, List<String> nameClassVariables, List<String> nameFeatureVariables, BNParameterLearningAlgorithm bnPLA, CTBNParameterLearningAlgorithm ctbnPLA, String initialStructure, double sigPC, double sigTimeTransitionHypothesis, double sigStateToStateTransitionHypothesis, List<Long> seeds, String nameGeneratedFile)
      Initialises the writer. This constructor was thought to report results of score-based and constraint-based algorithms.
      Parameters:
      nameDatasets - dataset names
      nameModels - model names
      nameClassVariables - names of class variables
      nameFeatureVariables - names of feature variables
      bnPLA - parameter learning algorithm for a Bayesian network
      ctbnPLA - parameter learning algorithm for a continuous-time Bayesian network
      initialStructure - name for the initial structure
      sigPC - significance level for the PC algorithms
      sigTimeTransitionHypothesis - significance level used for the null time to transition hypothesis
      sigStateToStateTransitionHypothesis - significance level used for the null state-to-state transition hypothesis
      seeds - list with seeds that are used to shuffle the datasets. It is assumed that each dataset will be evaluated with each of those seeds
      nameGeneratedFile - name of the generated Excel file

  • Method Details

    • close

      public void close()
      Description copied from class: MetricsWriter
      Closes the writer.
      Overrides:
      close in class MetricsWriter

    • write

      public void write(Map<String,Double> results)
      Description copied from class: MetricsWriter
      Writes the given results. The results are provided as a Map where ke evaluation metric names and values represent their corresponding scores.
      Specified by:
      write in class MetricsWriter
      Parameters:
      results - a Map with the results of the evaluation metrics