Make setters on ModelParameters protected, setN() and setD() methods …

…are called on fit() and the parameters are now part of BaseModelParameters, Rename print_r() to var_export().

pom.xml

@@ -128,6 +128,12 @@
                     <target>${java-version}</target>
                     <showDeprecation>true</showDeprecation>
                     <encoding>UTF-8</encoding>
+                    <compilerArgs>
+                        <!--
+                        <arg>-verbose</arg>
+                        <arg>-Xlint:all,-options,-path</arg>
+                        -->
+                    </compilerArgs>
                 </configuration>
             </plugin>
             <plugin>

src/main/java/com/datumbox/applications/datamodeling/Modeler.java

@@ -37,7 +37,7 @@ public class Modeler extends BaseWrapper<Modeler.ModelParameters, Modeler.Traini
      */
     public static class ModelParameters extends BaseWrapper.ModelParameters {
 
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 

src/main/java/com/datumbox/applications/nlp/TextClassifier.java

@@ -47,7 +47,7 @@ public class TextClassifier extends BaseWrapper<TextClassifier.ModelParameters,
      */
     public static class ModelParameters extends BaseWrapper.ModelParameters {
 
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 
@@ -119,24 +119,6 @@ public TextClassifier(String dbName, DatabaseConfiguration dbConf) {
         super(dbName, dbConf, TextClassifier.ModelParameters.class, TextClassifier.TrainingParameters.class);
     }
 
-    /**
-     * Trains a Machine Learning model using the provided training data.
-     * 
-     * @param trainingData
-     * @param trainingParameters 
-     */
-    @Override
-    public void fit(Dataset trainingData, TrainingParameters trainingParameters) { 
-        logger.info("fit()");
-
-        initializeTrainingConfiguration(trainingParameters);
-
-        _fit(trainingData);
-
-        //store database
-        knowledgeBase.save();
-    }
-
     /**
      * Trains a Machine Learning model using the provided dataset files. The data
      * map should have as keys the names of each class and as values the URIs

src/main/java/com/datumbox/common/utilities/PHPfunctions.java

@@ -259,7 +259,7 @@ public static <T> void shuffle(T[] array) {
      * @param object
      * @return 
      */
-    public static <T> String print_r(T object) {
+    public static <T> String var_export(T object) {
         return ToStringBuilder.reflectionToString(object);
     }
 

src/main/java/com/datumbox/framework/machinelearning/classification/BernoulliNaiveBayes.java

@@ -51,11 +51,11 @@ public static class ModelParameters extends BaseNaiveBayes.ModelParameters {
         private Map<Object, Double> sumOfLog1minusProb; //the Sum Of Log(1-prob) for each class. This is used to optimize the speed of validation. Instead of looping through all the keywords by having this Sum we are able to loop only through the features of the observation
 
         /**
-         * Public constructor which accepts as argument the DatabaseConnector.
+         * Protected constructor which accepts as argument the DatabaseConnector.
          * 
          * @param dbc 
          */
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 
@@ -73,7 +73,7 @@ public Map<Object, Double> getSumOfLog1minusProb() {
          * 
          * @param sumOfLog1minusProb 
          */
-        public void setSumOfLog1minusProb(Map<Object, Double> sumOfLog1minusProb) {
+        protected void setSumOfLog1minusProb(Map<Object, Double> sumOfLog1minusProb) {
             this.sumOfLog1minusProb = sumOfLog1minusProb;
         }
     } 
@@ -186,22 +186,18 @@ protected void predictDataset(Dataset newData) {
 
     @Override
     protected void _fit(Dataset trainingData) {
+        ModelParameters modelParameters = knowledgeBase.getModelParameters();
+        int n = modelParameters.getN();
+        int d = modelParameters.getD();
+
         knowledgeBase.getTrainingParameters().setMultiProbabilityWeighted(false);
 
-        ModelParameters modelParameters = knowledgeBase.getModelParameters();
 
         Map<List<Object>, Double> likelihoods = modelParameters.getLogLikelihoods();
         Map<Object, Double> logPriors = modelParameters.getLogPriors();
         Set<Object> classesSet = modelParameters.getClasses();
         Map<Object, Double> sumOfLog1minusProb = modelParameters.getSumOfLog1minusProb();
 
-        int n = trainingData.getRecordNumber();
-        int d = trainingData.getVariableNumber();
-
-        //initialization
-        modelParameters.setN(n);
-        modelParameters.setD(d);
-
 
         //calculate first statistics about the classes
         AssociativeArray totalFeatureOccurrencesForEachClass = new AssociativeArray(); 
@@ -261,10 +257,6 @@ protected void _fit(Dataset trainingData) {
 
         }
 
-        int c = classesSet.size();
-        modelParameters.setC(c);
-
-
         //calculate prior log probabilities
         for(Map.Entry<Object, Double> entry : logPriors.entrySet()) {
             Object theClass = entry.getKey();

src/main/java/com/datumbox/framework/machinelearning/classification/BinarizedNaiveBayes.java

@@ -40,11 +40,11 @@ public class BinarizedNaiveBayes extends BaseNaiveBayes<BinarizedNaiveBayes.Mode
     public static class ModelParameters extends BaseNaiveBayes.ModelParameters {
 
         /**
-         * Public constructor which accepts as argument the DatabaseConnector.
+         * Protected constructor which accepts as argument the DatabaseConnector.
          * 
          * @param dbc 
          */
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 

src/main/java/com/datumbox/framework/machinelearning/classification/MaximumEntropy.java

@@ -54,11 +54,11 @@ public static class ModelParameters extends BaseMLclassifier.ModelParameters {
         private Map<List<Object>, Double> lambdas; //the lambda parameters of the model
 
         /**
-         * Public constructor which accepts as argument the DatabaseConnector.
+         * Protected constructor which accepts as argument the DatabaseConnector.
          * 
          * @param dbc 
          */
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 
@@ -76,7 +76,7 @@ public Map<List<Object>, Double> getLambdas() {
          * 
          * @param lambdas 
          */
-        public void setLambdas(Map<List<Object>, Double> lambdas) {
+        protected void setLambdas(Map<List<Object>, Double> lambdas) {
             this.lambdas = lambdas;
         }
 
@@ -151,14 +151,7 @@ protected void predictDataset(Dataset newData) {
     @SuppressWarnings("unchecked")
     protected void _fit(Dataset trainingData) {
         ModelParameters modelParameters = knowledgeBase.getModelParameters();
-
-        int n = trainingData.getRecordNumber();
-        int d = trainingData.getVariableNumber();
-
-
-        //initialization
-        modelParameters.setN(n);
-        modelParameters.setD(d);
+        int n = modelParameters.getN();
 
 
         Map<List<Object>, Double> lambdas = modelParameters.getLambdas();
@@ -171,10 +164,6 @@ protected void _fit(Dataset trainingData) {
 
             classesSet.add(theClass); 
         }
-
-        int c = classesSet.size();
-        modelParameters.setC(c);
-
 
         //create a temporary map for the observed probabilities in training set
         DatabaseConnector dbc = knowledgeBase.getDbc();

src/main/java/com/datumbox/framework/machinelearning/classification/MultinomialNaiveBayes.java

@@ -38,11 +38,11 @@ public class MultinomialNaiveBayes extends BaseNaiveBayes<MultinomialNaiveBayes.
     public static class ModelParameters extends BaseNaiveBayes.ModelParameters {
 
         /**
-         * Public constructor which accepts as argument the DatabaseConnector.
+         * Protected constructor which accepts as argument the DatabaseConnector.
          * 
          * @param dbc 
          */
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 

src/main/java/com/datumbox/framework/machinelearning/classification/OrdinalRegression.java

@@ -67,11 +67,11 @@ public static class ModelParameters extends BaseMLclassifier.ModelParameters {
         private Map<Object, Double> thitas; 
 
         /**
-         * Public constructor which accepts as argument the DatabaseConnector.
+         * Protected constructor which accepts as argument the DatabaseConnector.
          * 
          * @param dbc 
          */
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 
@@ -89,7 +89,7 @@ public Map<Object, Double> getWeights() {
          * 
          * @param weights 
          */
-        public void setWeights(Map<Object, Double> weights) {
+        protected void setWeights(Map<Object, Double> weights) {
             this.weights = weights;
         }
 
@@ -107,7 +107,7 @@ public Map<Object, Double> getThitas() {
          * 
          * @param thitas 
          */
-        public void setThitas(Map<Object, Double> thitas) {
+        protected void setThitas(Map<Object, Double> thitas) {
             this.thitas = thitas;
         }
 
@@ -240,17 +240,9 @@ protected void predictDataset(Dataset newData) {
     @Override
     @SuppressWarnings("unchecked")
     protected void _fit(Dataset trainingData) {
-
-        int n = trainingData.getRecordNumber();
-        int d = trainingData.getVariableNumber();
-
         ModelParameters modelParameters = knowledgeBase.getModelParameters();
         TrainingParameters trainingParameters = knowledgeBase.getTrainingParameters();
-
-        //initialization
-        modelParameters.setN(n);
-        modelParameters.setD(d);
-
+
         Map<Object, Double> weights = modelParameters.getWeights();
         Map<Object, Double> thitas = modelParameters.getThitas();
 
@@ -265,9 +257,6 @@ protected void _fit(Dataset trainingData) {
         Set<Object> classesSet = modelParameters.getClasses();
         classesSet.addAll(sortedClasses);
 
-        int c = classesSet.size();
-        modelParameters.setC(c);
-
         //we initialize the weights and thitas to zero
         for(Object feature: trainingData.getXDataTypes().keySet()) {
             weights.put(feature, 0.0);

src/main/java/com/datumbox/framework/machinelearning/classification/SoftMaxRegression.java

@@ -54,11 +54,11 @@ public static class ModelParameters extends BaseMLclassifier.ModelParameters {
         private Map<List<Object>, Double> thitas; //the thita parameters of the model
 
         /**
-         * Public constructor which accepts as argument the DatabaseConnector.
+         * Protected constructor which accepts as argument the DatabaseConnector.
          * 
          * @param dbc 
          */
-        public ModelParameters(DatabaseConnector dbc) {
+        protected ModelParameters(DatabaseConnector dbc) {
             super(dbc);
         }
 
@@ -76,7 +76,7 @@ public Map<List<Object>, Double> getThitas() {
          * 
          * @param thitas 
          */
-        public void setThitas(Map<List<Object>, Double> thitas) {
+        protected void setThitas(Map<List<Object>, Double> thitas) {
             this.thitas = thitas;
         }
     } 
@@ -209,17 +209,9 @@ protected void predictDataset(Dataset newData) {
     @Override
     @SuppressWarnings("unchecked")
     protected void _fit(Dataset trainingData) {
-
-        int n = trainingData.getRecordNumber();
-        int d = trainingData.getVariableNumber();
-
         ModelParameters modelParameters = knowledgeBase.getModelParameters();
         TrainingParameters trainingParameters = knowledgeBase.getTrainingParameters();
 
-        //initialization
-        modelParameters.setN(n);
-        modelParameters.setD(d);
-
 
         Map<List<Object>, Double> thitas = modelParameters.getThitas();
         Set<Object> classesSet = modelParameters.getClasses();
@@ -231,9 +223,6 @@ protected void _fit(Dataset trainingData) {
 
             classesSet.add(theClass); 
         }
-
-        int c = classesSet.size();
-        modelParameters.setC(c);
 
         //we initialize the thitas to zero for all features and all classes compinations
         for(Object theClass : classesSet) {