make public to use v2 interface

Author: Ben Auffarth

src/main/java/org/lightgbm/predict4j/v2/Boosting.java

@@ -18,7 +18,7 @@ public abstract class Boosting implements Serializable {
     private static final Logger logger = LoggerFactory.getLogger(Boosting.class);
     private static final long serialVersionUID = -3370589073161617590L;
 
-    static Boosting createBoosting(String filename) throws FileNotFoundException, IOException {
+    public static Boosting createBoosting(String filename) throws IOException {
         String type = getBoostingTypeFromModelFile(filename);
         Boosting boosting = null;
         if (type.equals("tree")) {
@@ -30,7 +30,7 @@ static Boosting createBoosting(String filename) throws FileNotFoundException, IO
         return boosting;
     }
 
-    static Boosting createBoosting(String type, String filename) throws FileNotFoundException, IOException {
+    public static Boosting createBoosting(String type, String filename) throws FileNotFoundException, IOException {
         if (filename == null || filename.length() == 0) {
             if (type.equals("gbdt")) {
                 return new GBDT();
@@ -62,7 +62,7 @@ static Boosting createBoosting(String type, String filename) throws FileNotFound
         }
     }
 
-    static boolean loadFileToBoosting(Boosting boosting, String filename) throws FileNotFoundException, IOException {
+    public static boolean loadFileToBoosting(Boosting boosting, String filename) throws IOException {
         if (boosting != null) {
             StringBuilder sb = new StringBuilder();
             List<String> lines = IOUtils.readLines(new FileInputStream(filename));
@@ -76,28 +76,28 @@ static boolean loadFileToBoosting(Boosting boosting, String filename) throws Fil
         return true;
     }
 
-    static String getBoostingTypeFromModelFile(String filename) throws FileNotFoundException, IOException {
+    public static String getBoostingTypeFromModelFile(String filename) throws IOException {
         List<String> lines = IOUtils.readLines(new FileInputStream(filename));
         return lines.get(0);
     }
 
-    abstract boolean loadModelFromString(String modelStr);
+    abstract public boolean loadModelFromString(String modelStr);
 
-    abstract boolean needAccuratePrediction();
+    abstract public boolean needAccuratePrediction();
 
-    abstract int numberOfClasses();
+    abstract public int numberOfClasses();
 
-    abstract void initPredict(int num_iteration);
+    abstract public void initPredict(int num_iteration);
 
-    abstract int numPredictOneRow(int num_iteration, boolean is_pred_leaf);
+    abstract public int numPredictOneRow(int num_iteration, boolean is_pred_leaf);
 
-    abstract int getCurrentIteration();
+    abstract public int getCurrentIteration();
 
-    abstract int maxFeatureIdx();
+    abstract public int maxFeatureIdx();
 
-    abstract List<Double> predictLeafIndex(SparseVector vector);
+    abstract public List<Double> predictLeafIndex(SparseVector vector);
 
-    abstract List<Double> predictRaw(SparseVector vector, PredictionEarlyStopInstance early_stop);
+    abstract public List<Double> predictRaw(SparseVector vector, PredictionEarlyStopInstance early_stop);
 
-    abstract List<Double> predict(SparseVector vector, PredictionEarlyStopInstance early_stop);
+    abstract public List<Double> predict(SparseVector vector, PredictionEarlyStopInstance early_stop);
 }

src/main/java/org/lightgbm/predict4j/v2/GBDT.java

@@ -34,7 +34,8 @@ public class GBDT extends Boosting {
     /* ! \brief current iteration */
     int iter_;
 
-    boolean loadModelFromString(String model_str) {
+    @Override
+    public boolean loadModelFromString(String model_str) {
         // use serialized string to restore this object
         models_.clear();
         String[] lines = model_str.split("\n");
@@ -134,26 +135,26 @@ boolean loadModelFromString(String model_str) {
         return true;
     }
 
-    boolean needAccuratePrediction() {
+    public boolean needAccuratePrediction() {
         if (objective_function_ == null) {
             return true;
         } else {
             return objective_function_.needAccuratePrediction();
         }
     }
 
-    int numberOfClasses() {
+    public int numberOfClasses() {
         return num_class_;
     }
 
-    void initPredict(int num_iteration) {
+    public void initPredict(int num_iteration) {
         num_iteration_for_pred_ = models_.size() / num_tree_per_iteration_;
         if (num_iteration > 0) {
             num_iteration_for_pred_ = Math.min(num_iteration + (boost_from_average_ ? 1 : 0), num_iteration_for_pred_);
         }
     }
 
-    int numPredictOneRow(int num_iteration, boolean is_pred_leaf) {
+    public int numPredictOneRow(int num_iteration, boolean is_pred_leaf) {
         int num_preb_in_one_row = num_class_;
         if (is_pred_leaf) {
             int max_iteration = getCurrentIteration();
@@ -166,15 +167,15 @@ int numPredictOneRow(int num_iteration, boolean is_pred_leaf) {
         return num_preb_in_one_row;
     }
 
-    int getCurrentIteration() {
+    public int getCurrentIteration() {
         return models_.size() / num_tree_per_iteration_;
     }
 
-    int maxFeatureIdx() {
+    public int maxFeatureIdx() {
         return max_feature_idx_;
     }
 
-    List<Double> predictLeafIndex(SparseVector vector) {
+    public List<Double> predictLeafIndex(SparseVector vector) {
         List<Double> outputs=new ArrayList<>();
         for (int i = 0; i < num_iteration_for_pred_; ++i) {
             for (int j = 0; j < num_class_; ++j) {
@@ -184,7 +185,7 @@ List<Double> predictLeafIndex(SparseVector vector) {
         return outputs;
     }
 
-    List<Double> predictRaw(SparseVector features, PredictionEarlyStopInstance early_stop) {
+    public List<Double> predictRaw(SparseVector features, PredictionEarlyStopInstance early_stop) {
         double[] output = new double[num_class_];
         int early_stop_round_counter = 0;
         for (int i = 0; i < num_iteration_for_pred_; ++i) {
@@ -204,7 +205,7 @@ List<Double> predictRaw(SparseVector features, PredictionEarlyStopInstance early
         return ret;
     }
 
-    List<Double> predict(SparseVector features, PredictionEarlyStopInstance early_stop) {
+    public List<Double> predict(SparseVector features, PredictionEarlyStopInstance early_stop) {
         List<Double> ret=predictRaw(features, early_stop);;
         if (objective_function_ != null) {
             double[]output=new double[ret.size()];