Bug fixes to deep learning module and TDisplayFX

src/dataPlotsFX/TDGraphParametersFX.java

@@ -80,6 +80,18 @@ public class TDGraphParametersFX implements Serializable, Cloneable, ManagedPara
 	 */
 	public String plotFillS = "white";
 	
+	/**
+	 * Show the hide panel on the left on start up. Note only used when first opening 
+	 * saved settings or saving settings. 
+	 */
+	public boolean showHidePaneLeft = true;
+	
+	/**
+	 * Show the hide panel on the right on start up. Note only used when first opening 
+	 * saved settings or saving settings. 
+	 */
+	public boolean showHidePaneRight = false;
+
 
 	/* (non-Javadoc)
 	 * @see java.lang.Object#clone()

src/dataPlotsFX/layout/TDGraphFX.java

@@ -350,8 +350,8 @@ public class TDGraphFX extends PamBorderPane {
 		setOverlayColour(LIGHT_TD_DISPLAY);
 		
 		
-		//show the left hiding pane byu default. 
-		stackPane.getLeftHidingPane().showHidePane(true);
+//		//show the left hiding pane byu default. 
+//		stackPane.getLeftHidingPane().showHidePane(true);
 	}
 
 	/**
@@ -1699,8 +1699,7 @@ public class TDGraphFX extends PamBorderPane {
 
 	/**
 	 * A bit different to the standard getter in that this only gets called just
-	 * before the configuration is serialized into the .psf. It's time to pull any
-	 * configuration information out about every line drawn on this boomin' thing !
+	 * before the configuration is serialized into the .psfx.
 	 * 
 	 * @return graph parameters ready to serialised.
 	 */
@@ -1733,6 +1732,11 @@ public class TDGraphFX extends PamBorderPane {
 					graphParameters.setScaleInfoData(scaleInfo.getDataTypeInfo(), scaleInfo.getScaleInfoData());
 				}
 			}
+			
+			//Finally save whether the hiding panels are open or not. 
+			graphParameters.showHidePaneLeft = stackPane.getLeftHidingPane().isShowing();
+			graphParameters.showHidePaneRight = stackPane.getRightHidingPane().isShowing();
+
 		}
 
 		return graphParameters;
@@ -1817,6 +1821,11 @@ public class TDGraphFX extends PamBorderPane {
 		tdAxisSelPane.selectAxisType();
 		setAxisName(graphParameters.currentDataType);
 		
+		//Open hide panes if needed. 
+		//Finally save whether the hiding panels are open or not. 
+		stackPane.getLeftHidingPane().showHidePane(graphParameters.showHidePaneLeft);
+		stackPane.getRightHidingPane().showHidePane(graphParameters.showHidePaneRight);
+
 	}
 
 	/**

src/dataPlotsFX/rawClipDataPlot/RawClipDataInfo.java

@@ -2,6 +2,7 @@ package dataPlotsFX.rawClipDataPlot;
 
 
 import java.awt.geom.Path2D;
+import java.io.Serializable;
 import java.util.ArrayList;
 import java.util.Arrays;
 
@@ -14,7 +15,7 @@ import PamguardMVC.PamDataUnit;
 import PamguardMVC.RawDataHolder;
 import clipgenerator.ClipSpectrogram;
 import dataPlotsFX.TDSymbolChooserFX;
-import dataPlotsFX.clickPlotFX.ClickSymbolChooserFX;
+import dataPlotsFX.clickPlotFX.ClickDisplayParams;
 import dataPlotsFX.data.TDDataProviderFX;
 import dataPlotsFX.data.TDScaleInfo;
 import dataPlotsFX.data.generic.GenericDataPlotInfo;
@@ -256,5 +257,35 @@ public class RawClipDataInfo extends GenericDataPlotInfo {
 		
 	}
 	
+	/* (non-Javadoc)
+	 * @see dataPlots.data.TDDataInfo#getStoredSettings()
+	 */
+	@Override
+	public Serializable getStoredSettings() {
+		return rawClipParams;
+	}
+
+	/* (non-Javadoc)
+	 * @see dataPlots.data.TDDataInfo#setStoredSettings(java.io.Serializable)
+	 */
+	@Override
+	public boolean setStoredSettings(Serializable storedSettings) {
+		if (RawClipParams.class.isAssignableFrom(storedSettings.getClass())) {
+			rawClipParams = (RawClipParams) storedSettings;
+			updateSettings();
+			return true;
+		}
+		return false;
+	}
+
+	/**
+	 * Called whenever settings are updated. 
+	 */
+	private void updateSettings() {
+		// TODO Auto-generated method stub
+		
+	}
+
+
 
 }

src/rawDeepLearningClassifier/RawDLParams.java

@@ -38,6 +38,13 @@ public class RawDLParams implements Serializable, Cloneable {
 	 */
 	public GroupedSourceParameters groupedSourceParams = new GroupedSourceParameters();
 	
+	/**
+	 * True to enable segmentation. If segmentation is disabled then the raw waveform from 
+	 * a data unit is passed directly to the model. Note that this is not an option for raw sound
+	 * data. 
+	 */
+	public boolean enableSegmentation = true;
+
 	/**
 	 * The number of raw samples to send to the classifier.
 	 */
@@ -85,7 +92,7 @@ public class RawDLParams implements Serializable, Cloneable {
 	 * different class names. If we change model then the class names may change.
 	 * Previously annotated data will then be messed up. But, in a giant dataset
 	 * that may be an issue. Perhaps users wish to run a new model on some chunk of
-	 * data without messing up all the other classified detectionS which have used
+	 * data without messing up all the other classified detections which have used
 	 * that module. So store the data in binary files? That is super inefficient as
 	 * the same string is stored many times. So instead store a short which
 	 * identifies the string that sits in this table. Everytime a new model is added

src/rawDeepLearningClassifier/dataSelector/DLPredictonPane.java

@@ -141,12 +141,12 @@ public class DLPredictonPane extends DynamicSettingsPane<DLPredictionFilterParam
 	@Override
 	public DLPredictionFilterParams getParams(DLPredictionFilterParams currParams) {
 		
+		if (classPanes==null) return currParams;
 		for (int i=0; i<classPanes.length ; i++) {
 			currParams.classSelect[i] = classPanes[i].enable.isSelected();
 			currParams.minClassPredicton[i] = classPanes[i].slider.getValue();
 		}
 		
-		
 		return currParams;
 	}
 

src/rawDeepLearningClassifier/dlClassification/delphinID/DelphinIDUtils.java

@@ -382,6 +382,7 @@ public class DelphinIDUtils {
 		//		String modelPath = "/Users/au671271/Library/CloudStorage/Dropbox/PAMGuard_dev/Deep_Learning/delphinID/testencounter415/whistle_model_2/whistle_4s_415.zip";
 		String modelPath = "D:/Dropbox/PAMGuard_dev/Deep_Learning/delphinID/testencounter415/whistle_model_2/whistle_4s_415.zip";
 
+		//line widths in pixels
 		double[] lineWidths = new double[] {6, 7, 10, 15, 20};
 
 		for (double lineWidth:lineWidths) {

src/rawDeepLearningClassifier/segmenter/SegmenterProcess.java

@@ -486,9 +486,17 @@ public class SegmenterProcess extends PamProcess {
 
 		//pass the raw click data to the segmenter
 		for (int i=0;i<chans.length; i++) {
-			newRawData(pamDataUnit,
-					rawDataChunk[i], chans[i], true);
 			
+				if (dlControl.getDLParams().enableSegmentation) {
+					//segment the data unit into different chunks. 
+					newRawData(pamDataUnit,
+							rawDataChunk[i], chans[i], dlControl.getDLParams().rawSampleSize, dlControl.getDLParams().sampleHop, true);
+				}
+				else {
+					//send the whole data chunk to the deep learning unit
+					newRawData(pamDataUnit,
+							rawDataChunk[i], chans[i], 	rawDataChunk[i].length, rawDataChunk[i].length, true);
+				}
 			//the way that the newRawdata works is it waits for the next chunk and copies all relevant bits
 			//from previous chunks into segments. This is fine for continuous data but means that chunks of data
 			//don't get their last hop...
@@ -512,7 +520,7 @@ public class SegmenterProcess extends PamProcess {
 	 * @param iChan        - the channel that is being segmented
 	 */
 	public void newRawData(PamDataUnit unit, double[] rawDataChunk, int iChan) {
-		newRawData(unit, rawDataChunk, iChan, false); 
+		newRawData(unit, rawDataChunk, iChan, dlControl.getDLParams().rawSampleSize ,dlControl.getDLParams().sampleHop , false); 
 	}
 
 	/**
@@ -523,15 +531,17 @@ public class SegmenterProcess extends PamProcess {
 	 * 
 	 * @param unit         - the data unit which contains relevant metadata on time
 	 *                     etc.
-	 * @param rawDataChunk - the sound chunk to segment extracted from the data
+	 * @param rawDataChunk - the sound chunk extracted from the data
 	 *                     unit.
 	 * @param iChan        - the channel that is being segmented
+	 * @param rawSampleSize - the segment size in samples i.e. the size of the segmenting window. 
+	 * @param rawSampleHop - the segment hop in samples i.e. how far the window jumps for each segment. 
 	 * @param forceSave    - make sure that all data is passed into the buffers and
 	 *                     do not wait for the next data unit. This is used to make
 	 *                     sure that discrete chunks have their full number of
 	 *                     segments saved.
 	 */
-	public synchronized void newRawData(PamDataUnit unit, double[] rawDataChunk, int iChan, boolean forcesave) {
+	public synchronized void newRawData(PamDataUnit unit, double[] rawDataChunk, int iChan, int rawSampleSize, int rawSampleHop, boolean forcesave) {
 
 		long timeMilliseconds = unit.getTimeMilliseconds();
 		long startSampleTime = unit.getStartSample(); 
@@ -555,7 +565,7 @@ public class SegmenterProcess extends PamProcess {
 				if (currentRawChunks[i]==null) {
 					//create a new data unit - should only be called once after initial start.  
 					currentRawChunks[i] = new GroupedRawData(timeMilliseconds, getSourceParams().getGroupChannels(i), 
-							startSampleTime, dlControl.getDLParams().rawSampleSize, dlControl.getDLParams().rawSampleSize); 
+							startSampleTime, rawSampleSize, rawSampleSize); 
 					
 					currentRawChunks[i].setParentDataUnit(unit);; 
 				}
@@ -615,7 +625,7 @@ public class SegmenterProcess extends PamProcess {
 						
 						//segments which do not include any last zero padded segmen- zeros can confuse deep learning models so it may be better to keep use 
 						//this instead of zero padding end chunks. 
-						int nChunks = (int) Math.ceil((overFlow)/(double) dlControl.getDLParams().sampleHop); 
+						int nChunks = (int) Math.ceil((overFlow)/(double) rawSampleHop); 
 
 						nChunks = Math.max(nChunks, 1); //cannot be less than one (if forceSave is used then can be zero if no overflow)
 						nextRawChunks[i]=new GroupedRawData[nChunks]; 
@@ -638,11 +648,11 @@ public class SegmenterProcess extends PamProcess {
 
 							//go from current raw chunks tim millis to try and minimise compounding time errors. 
 							//							long timeMillis = (long) (currentRawChunks[i].getTimeMilliseconds() + j*(1000.*(dlControl.getDLParams().sampleHop)/this.getSampleRate())); 
-							long startSample = lastRawDataChunk.getStartSample() + dlControl.getDLParams().sampleHop; 
+							long startSample = lastRawDataChunk.getStartSample() + rawSampleHop; 
 							long timeMillis = this.absSamplesToMilliseconds(startSample); 
 
 							nextRawChunks[i][j] = new GroupedRawData(timeMillis, getSourceParams().getGroupChannels(i), 
-									startSample, dlControl.getDLParams().rawSampleSize, dlControl.getDLParams().rawSampleSize); 
+									startSample, rawSampleSize, rawSampleSize); 
 							nextRawChunks[i][j].setParentDataUnit(unit);
 
 						}

src/test/rawDeepLearningClassifier/ClickDLTest.java

@@ -0,0 +1,31 @@
+package test.rawDeepLearningClassifier;
+
+import org.junit.jupiter.api.Test;
+
+import rawDeepLearningClassifier.dlClassification.genericModel.GenericModelParams;
+import rawDeepLearningClassifier.dlClassification.genericModel.GenericModelWorker;
+
+public class ClickDLTest {
+
+	
+	@Test
+	public void clickDLTest() {
+		//relative paths to the resource folders.
+		System.out.println("*****Click classification Deep Learning*****"); 
+
+		//relative paths to the resource folders.
+		String relModelPath  =	"D:/Dropbox/PAMGuard_dev/Deep_Learning/click_classifier_Thomas/best_model/saved_model.pb";
+
+
+		GenericModelWorker genericModelWorker = new GenericModelWorker(); 
+
+		GenericModelParams genericModelParams = new GenericModelParams(); 
+
+		genericModelParams.modelPath = relModelPath;
+		
+		genericModelWorker.prepModel(genericModelParams, null);
+
+	
+	}
+	
+}

src/test/rawDeepLearningClassifier/PamZipDLClassifierTest.java

@@ -5,8 +5,8 @@ import org.junit.jupiter.api.Test;
 public class PamZipDLClassifierTest {
 	
 	/**
-	 * Test the koogu classifier and tests are working properly. This tests loading the koogu model and also using
-	 * functions in KooguWorker.
+	 * Test the koogu classifier and tests are working properly for a PAMGuard zip model - i.e. this is a very similar model to Koogu but zipped with a .zip 
+	 * filename instead of .kgu. 
 	 */
 	@Test
 	public void zipClassifierTest() {
@@ -21,4 +21,7 @@ public class PamZipDLClassifierTest {
 		KooguDLClassifierTest.runKooguClassifier( relModelPath,  relWavPath,  relMatPath);
 	}
 	
+	
+	
+
 }