Removing custom patch from AbstractBoostingBagging.

CHANGELOG.md

@@ -20,7 +20,9 @@ Version 0.8.0-SNAPSHOT - Build 20161221
 - Removed the AbstractWrapper and Modeler inherits directly from AbstractTrainer. 
 - Created a TrainableBundle to keep track of the Trainables of Modeler, AbstractBoostingBagging and StepwiseRegression.
 - Removed automatic save after fit, now save() must be called.
-- AbstractTrainer no longer stores a local copy of dbName.
+- AbstractTrainer no longer stores a local copy of dbName. The save method accepts a dbName.
+- The DatabaseConfiguration.getDBnameSeparator() method was removed.
+- The closeAndRename() is created in DatabaseConnectors and it's used by KnowledgeBase to saveAs the models.
 
 Version 0.7.1-SNAPSHOT - Build 20161217
 ---------------------------------------

TODO.txt

@@ -1,6 +1,7 @@
 CODE IMPROVEMENTS
 =================
 
+- Can we add all the files of a model in a single folder.
 - Can we make the two constructors of the Trainers to call a common constructor to eliminate duplicate code?
 
 - Support of better Transformers (Zscore, decouple boolean transforming from numeric etc).

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

@@ -35,6 +35,10 @@
  */
 public class Modeler extends AbstractTrainer<Modeler.ModelParameters, Modeler.TrainingParameters> implements Parallelizable {
 
+    private static final String DT_KEY = "dt";
+    private static final String FS_KEY = "fs";
+    private static final String ML_KEY = "ml";
+
     private TrainableBundle bundle = new TrainableBundle();
 
     /**
@@ -173,15 +177,15 @@ public void predict(Dataframe newData) {
         bundle.setParallelized(isParallelized());
 
         //run the pipeline
-        AbstractTransformer dataTransformer = (AbstractTransformer) bundle.get("dataTransformer");
+        AbstractTransformer dataTransformer = (AbstractTransformer) bundle.get(DT_KEY);
         if(dataTransformer != null) {
             dataTransformer.transform(newData);
         }
-        AbstractFeatureSelector featureSelector = (AbstractFeatureSelector) bundle.get("featureSelector");
+        AbstractFeatureSelector featureSelector = (AbstractFeatureSelector) bundle.get(FS_KEY);
         if(featureSelector != null) {
             featureSelector.transform(newData);
         }
-        AbstractModeler modeler = (AbstractModeler) bundle.get("modeler");
+        AbstractModeler modeler = (AbstractModeler) bundle.get(ML_KEY);
         modeler.predict(newData);
         if(dataTransformer != null) {
             dataTransformer.denormalize(newData);
@@ -193,7 +197,6 @@ public void predict(Dataframe newData) {
     protected void _fit(Dataframe trainingData) {
         TrainingParameters trainingParameters = knowledgeBase.getTrainingParameters();
         Configuration conf = knowledgeBase.getConf();
-        String dbName = knowledgeBase.getDbc().getDatabaseName();
 
         //reset previous entries on the bundle
         resetBundle();
@@ -203,19 +206,19 @@ protected void _fit(Dataframe trainingData) {
         AbstractTransformer dataTransformer = null;
         if(dtParams != null) {
             dataTransformer = MLBuilder.create(dtParams, conf);
-            bundle.put("dataTransformer", dataTransformer);
+            bundle.put(DT_KEY, dataTransformer);
         }
 
         AbstractFeatureSelector.AbstractTrainingParameters fsParams = trainingParameters.getFeatureSelectorTrainingParameters();
         AbstractFeatureSelector featureSelector = null;
         if(fsParams != null) {
             featureSelector = MLBuilder.create(fsParams, conf);
-            bundle.put("featureSelector", featureSelector);
+            bundle.put(FS_KEY, featureSelector);
         }
 
         AbstractModeler.AbstractTrainingParameters mlParams = trainingParameters.getModelerTrainingParameters();
         AbstractModeler modeler = MLBuilder.create(mlParams, conf);
-        bundle.put("modeler", modeler);
+        bundle.put(ML_KEY, modeler);
 
         //set the parallized flag to all algorithms
         bundle.setParallelized(isParallelized());
@@ -267,30 +270,30 @@ private void initBundle() {
         Configuration conf = knowledgeBase.getConf();
         String dbName = knowledgeBase.getDbc().getDatabaseName();
 
-        if(!bundle.containsKey("dataTransformer")) {
+        if(!bundle.containsKey(DT_KEY)) {
             AbstractTransformer.AbstractTrainingParameters dtParams = trainingParameters.getDataTransformerTrainingParameters();
 
             AbstractTransformer dataTransformer = null;
             if(dtParams != null) {
-                dataTransformer = MLBuilder.load(dtParams.getTClass(), dbName, conf);
+                dataTransformer = MLBuilder.load(dtParams.getTClass(), dbName + "_" + DT_KEY, conf);
             }
-            bundle.put("dataTransformer", dataTransformer);
+            bundle.put(DT_KEY, dataTransformer);
         }
 
-        if(!bundle.containsKey("featureSelector")) {
+        if(!bundle.containsKey(FS_KEY)) {
             AbstractFeatureSelector.AbstractTrainingParameters fsParams = trainingParameters.getFeatureSelectorTrainingParameters();
 
             AbstractFeatureSelector featureSelector = null;
             if(fsParams != null) {
-                featureSelector = MLBuilder.load(fsParams.getTClass(), dbName, conf);
+                featureSelector = MLBuilder.load(fsParams.getTClass(), dbName + "_" + FS_KEY, conf);
             }
-            bundle.put("featureSelector", featureSelector);
+            bundle.put(FS_KEY, featureSelector);
         }
 
-        if(!bundle.containsKey("modeler")) {
+        if(!bundle.containsKey(ML_KEY)) {
             AbstractModeler.AbstractTrainingParameters mlParams = trainingParameters.getModelerTrainingParameters();
 
-            bundle.put("modeler", MLBuilder.load(mlParams.getTClass(), dbName, conf));
+            bundle.put(ML_KEY, MLBuilder.load(mlParams.getTClass(), dbName + "_" + ML_KEY, conf));
         }
     }
 

datumbox-framework-core/src/main/java/com/datumbox/framework/core/machinelearning/common/abstracts/algorithms/AbstractBoostingBagging.java

@@ -167,20 +167,20 @@ protected void _predict(Dataframe newData) {
         //using the weak classifiers
         AssociativeArray classifierWeightsArray = new AssociativeArray();
         int totalWeakClassifiers = weakClassifierWeights.size();
-        for(int t=0;t<totalWeakClassifiers;++t) {
+        for(int i=0;i<totalWeakClassifiers;++i) {
 
-            AbstractClassifier mlclassifier = (AbstractClassifier) bundle.get(String.valueOf(t));
+            AbstractClassifier mlclassifier = (AbstractClassifier) bundle.get(DB_INDICATOR + i);
             mlclassifier.predict(newData);
 
-            classifierWeightsArray.put(t, weakClassifierWeights.get(t));
+            classifierWeightsArray.put(i, weakClassifierWeights.get(i));
 
             for(Map.Entry<Integer, Record> e : newData.entries()) {
                 Integer rId = e.getKey();
                 Record r = e.getValue();
                 AssociativeArray classProbabilities = r.getYPredictedProbabilities();
 
                 DataTable2D rDecisions = tmp_recordDecisions.get(rId);
-                rDecisions.put(t, classProbabilities);
+                rDecisions.put(i, classProbabilities);
 
                 tmp_recordDecisions.put(rId, rDecisions); //WARNING: Do not remove this! We must put it back to the Map to persist it on Disk-backed maps
             }
@@ -258,7 +258,7 @@ protected void _fit(Dataframe trainingData) {
                 mlclassifier.close();
             }
             else {
-                bundle.put(String.valueOf(i), mlclassifier);
+                bundle.put(DB_INDICATOR + i, mlclassifier);
             }
 
             if(status==Status.STOP) {
@@ -319,13 +319,7 @@ protected enum Status {
     public void save(String dbName) {
         initBundle();
         String knowledgeBaseName = createKnowledgeBaseName(dbName);
-        //bundle.save(knowledgeBaseName);
-        for(String i : bundle.keySet()) { //TODO: remove this custom case if possible. try adding the key in the name.
-            Trainable t = bundle.get(i);
-            if(t != null) {
-                t.save(knowledgeBaseName + "_" + DB_INDICATOR + i);
-            }
-        }
+        bundle.save(knowledgeBaseName);
         super.save(dbName);
     }
 
@@ -359,8 +353,9 @@ private void initBundle() {
         Class<AbstractClassifier> weakClassifierClass = trainingParameters.getWeakClassifierTrainingParameters().getTClass();
         int totalWeakClassifiers = Math.min(modelParameters.getWeakClassifierWeights().size(), trainingParameters.getMaxWeakClassifiers());
         for(int i=0;i<totalWeakClassifiers;i++) {
-            if (!bundle.containsKey(String.valueOf(i))) {
-                bundle.put(String.valueOf(i), MLBuilder.load(weakClassifierClass, dbc.getDatabaseName() + "_" + DB_INDICATOR + i, conf));
+            String key = DB_INDICATOR + i;
+            if (!bundle.containsKey(key)) {
+                bundle.put(key, MLBuilder.load(weakClassifierClass, dbc.getDatabaseName() + "_" + key, conf));
             }
         }
     }

datumbox-framework-core/src/main/java/com/datumbox/framework/core/machinelearning/common/dataobjects/TrainableBundle.java

@@ -89,9 +89,10 @@ public void setParallelized(boolean parallelized) {
 
     /** {@inheritDoc} */
     public void save(String dbName) {
-        for(Trainable t : bundle.values()) {
+        for(Map.Entry<String, Trainable> e : bundle.entrySet()) {
+            Trainable t = e.getValue();
             if(t != null) {
-                t.save(dbName);
+                t.save(dbName + "_" + e.getKey());
             }
         }
     }

datumbox-framework-core/src/main/java/com/datumbox/framework/core/machinelearning/regression/StepwiseRegression.java

@@ -39,6 +39,8 @@
  */
 public class StepwiseRegression extends AbstractRegressor<StepwiseRegression.ModelParameters, StepwiseRegression.TrainingParameters>  {
 
+    private static final String REG_KEY = "reg";
+
     private TrainableBundle bundle = new TrainableBundle();
 
     /** {@inheritDoc} */
@@ -158,7 +160,7 @@ protected void _predict(Dataframe newData) {
         initBundle();
 
         //run the pipeline
-        AbstractRegressor mlregressor = (AbstractRegressor) bundle.get("mlregressor");
+        AbstractRegressor mlregressor = (AbstractRegressor) bundle.get(REG_KEY);
         mlregressor.predict(newData);
     }
 
@@ -167,7 +169,6 @@ protected void _predict(Dataframe newData) {
     protected void _fit(Dataframe trainingData) {
         TrainingParameters trainingParameters = knowledgeBase.getTrainingParameters();
         Configuration conf = knowledgeBase.getConf();
-        String dbName = knowledgeBase.getDbc().getDatabaseName();
 
         //reset previous entries on the bundle
         resetBundle();
@@ -217,7 +218,7 @@ protected void _fit(Dataframe trainingData) {
                 conf
         );
         mlregressor.fit(copiedTrainingData);
-        bundle.put("mlregressor", mlregressor);
+        bundle.put(REG_KEY, mlregressor);
 
         copiedTrainingData.delete();
     }
@@ -256,10 +257,10 @@ private void initBundle() {
         Configuration conf = knowledgeBase.getConf();
         String dbName = knowledgeBase.getDbc().getDatabaseName();
 
-        if(!bundle.containsKey("mlregressor")) {
+        if(!bundle.containsKey(REG_KEY)) {
             AbstractTrainingParameters mlParams = trainingParameters.getRegressionTrainingParameters();
 
-            bundle.put("mlregressor", MLBuilder.load(mlParams.getTClass(), dbName, conf));
+            bundle.put(REG_KEY, MLBuilder.load(mlParams.getTClass(), dbName + "_" + REG_KEY, conf));
         }
     }