Fixed main demo scripts to be compatible with Matlab (fixed printf -> printf, rows -> size(x, 1), and not specifically converting matlab arrays to java and back)

demos/octave/example1TeBinaryData.m

@@ -16,10 +16,10 @@ teCalc=javaObject('infodynamics.measures.discrete.ApparentTransferEntropyCalcula
 teCalc.initialise();
 % Since we have simple arrays of doubles, we can directly pass these in:
 teCalc.addObservations(destArray, sourceArray);
-printf("For copied source, result should be close to 1 bit : ");
+fprintf('For copied source, result should be close to 1 bit : ');
 result = teCalc.computeAverageLocalOfObservations()
 teCalc.initialise();
 teCalc.addObservations(destArray, sourceArray2);
-printf("For random source, result should be close to 0 bits: ");
+fprintf('For random source, result should be close to 0 bits: ');
 result2 = teCalc.computeAverageLocalOfObservations()
 

demos/octave/example2TeMultidimBinaryData.m

@@ -25,6 +25,6 @@ teCalc=javaObject('infodynamics.measures.discrete.ApparentTransferEntropyCalcula
 teCalc.initialise();
 % Add observations of transfer across one cell to the right per time step:
 teCalc.addObservations(twoDTimeSeriesJavaInt, 1);
-printf("The result should be close to 1 bit here, since we are executing copy operations of what is effectively a random bit to each cell here: ");
+fprintf('The result should be close to 1 bit here, since we are executing copy operations of what is effectively a random bit to each cell here: ');
 result2D = teCalc.computeAverageLocalOfObservations()
 

demos/octave/example3TeContinuousDataKernel.m

@@ -13,17 +13,17 @@ destArray = [0; covariance*sourceArray(1:numObservations-1) + (1-covariance)*nor
 sourceArray2=normrnd(0, 1, numObservations, 1); % Uncorrelated source
 % Create a TE calculator and run it:
 teCalc=javaObject('infodynamics.measures.continuous.kernel.TransferEntropyCalculatorKernel');
-teCalc.setProperty("NORMALISE_PROP_NAME", "true"); % Normalise the individual variables
+teCalc.setProperty('NORMALISE_PROP_NAME', 'true'); % Normalise the individual variables
 teCalc.initialise(1, 0.5); % Use history length 1 (Schreiber k=1), kernel width of 0.5 normalised units
 teCalc.setObservations(sourceArray, destArray);
 % For copied source, should give something close to 1 bit:
 result = teCalc.computeAverageLocalOfObservations();
-printf("TE result %.4f bits; expected to be close to %.4f bits for these correlated Gaussians but biased upwards\n", \
+fprintf('TE result %.4f bits; expected to be close to %.4f bits for these correlated Gaussians but biased upwards\n', ...
     result, log(1/(1-covariance^2))/log(2));
 teCalc.initialise(); % Initialise leaving the parameters the same
 teCalc.setObservations(sourceArray2, destArray);
 % For random source, it should give something close to 0 bits
 result2 = teCalc.computeAverageLocalOfObservations();
-printf("TE result %.4f bits; expected to be close to 0 bits for uncorrelated Gaussians but will be biased upwards\n", \
+fprintf('TE result %.4f bits; expected to be close to 0 bits for uncorrelated Gaussians but will be biased upwards\n', ...
     result2);
 

demos/octave/example4TeContinuousDataKraskov.m

@@ -14,7 +14,7 @@ sourceArray2=normrnd(0, 1, numObservations, 1); % Uncorrelated source
 % Create a TE calculator and run it:
 teCalc=javaObject('infodynamics.measures.continuous.kraskov.TransferEntropyCalculatorKraskov');
 teCalc.initialise(1); % Use history length 1 (Schreiber k=1)
-teCalc.setProperty("k", "4"); % Use Kraskov parameter K=4 for 4 nearest points
+teCalc.setProperty('k', '4'); % Use Kraskov parameter K=4 for 4 nearest points
 % Perform calculation with correlated source:
 teCalc.setObservations(sourceArray, destArray);
 result = teCalc.computeAverageLocalOfObservations();
@@ -22,12 +22,12 @@ result = teCalc.computeAverageLocalOfObservations();
 %  data is a set of random variables) - the result will be of the order
 %  of what we expect, but not exactly equal to it; in fact, there will
 %  be a large variance around it.
-printf("TE result %.4f nats; expected to be close to %.4f nats for these correlated Gaussians\n", \
+fprintf('TE result %.4f nats; expected to be close to %.4f nats for these correlated Gaussians\n', ...
     result, log(1/(1-covariance^2)));
 % Perform calculation with uncorrelated source:
 teCalc.initialise(); % Initialise leaving the parameters the same
 teCalc.setObservations(sourceArray2, destArray);
 result2 = teCalc.computeAverageLocalOfObservations();
-printf("TE result %.4f nats; expected to be close to 0 nats for these uncorrelated Gaussians\n", result2);
+fprintf('TE result %.4f nats; expected to be close to 0 nats for these uncorrelated Gaussians\n', result2);
 
 

demos/octave/javaMatrixToOctave.m

@@ -1,6 +1,6 @@
 % function octaveMatrix = javaMatrixToOctave(javaMatrix)
 %
-% Convert a java matrix (1 or 2D, double or int - but not Integer!!) to an octave matrix
+% Convert a java matrix (1 or 2D, double or int - but not Integer!!) to an octave or matlab matrix
 %
 % Octave-java doesn't seem to handle the conversion natively,
 %  so we either use org.octave.Matrix (built-in) to do it, or
@@ -16,10 +16,10 @@ function octaveMatrix = javaMatrixToOctave(javaMatrix, startRow, startCol, numRo
 		startCol = 1;
 	end
 	if (nargin < 4)
-		numRows = rows(javaMatrix);
+		numRows = size(javaMatrix, 1);
 	end
 	if (nargin < 5)
-		numCols = columns(javaMatrix);
+		numCols = size(javaMatrix, 2);
 	end
 
 	if (exist ('OCTAVE_VERSION', 'builtin'))
@@ -46,11 +46,12 @@ function octaveMatrix = javaMatrixToOctave(javaMatrix, startRow, startCol, numRo
 			end
 			return;
 		end
-		% else fall through to cell by cell conversion, as per for matlab
+	else
+		% Else we're in matlab, in which case the native java type can be handled, so return it directly:
+		octaveMatrix = javaMatrix;
 	end
 
-	% Else, either we encountered an error in the octave resizing,
-	%  or we were in Matlab all along. (If there's a fast way for matlab, tell me)
+	% Else, we encountered an error in the octave resizing, so fall through to element by element conversion:
 	
 	octaveMatrix = zeros(numRows, numCols);
 	for r = startRow:startRow+numRows-1

demos/octave/octaveToJavaDoubleArray.m

@@ -24,16 +24,9 @@ function jDoubleArray = octaveToJavaDoubleArray(octaveArray)
 			jDoubleArray(1) = octaveArray(1);
 		end
 	else
-		% We're in matlab:
+		% We're in matlab: the native matlab array can be passed to java as is:
 		
-		% Presumably there's a quick way to do this in matlab, but since I'm not on matlab, I don't know ...
-		% If someone knows or has tested something, please tell me and I'll include it here.
-		% In the meantime, we copy element by element
-		
-		jDoubleArray = javaArray('java.lang.Double', length(octaveArray));
-		for r = 1:length(octaveMatrix)
-			jDoubleArray(r) = octaveArray(r);
-		end
+		jDoubleArray = octaveArray;
 	end
 
 end

demos/octave/octaveToJavaDoubleMatrix.m

@@ -21,21 +21,9 @@ function jDoubleMatrix = octaveToJavaDoubleMatrix(octaveMatrix)
 			jDoubleMatrix(1, 1) = octaveMatrix(1);
 		end
 	else
-		% We're in matlab:
+		% We're in matlab: the native matlab 2D array can be passed to java as is:
 		
-		% Presumably there's a quick way to do this in matlab, but since I'm not on matlab, I don't know ...
-		% If someone knows or has tested something, please tell me and I'll include it here.
-		% In the meantime, we copy element by element
-		
-		jDoubleMatrix = javaArray('java.lang.Double', rows(octaveMatrix), columns(octaveMatrix));
-		for r = 1:rows(octaveMatrix)
-			% Slow but effective way:
-			for c = 1:columns(octaveMatrix)
-				jDoubleMatrix(r,c) = octaveMatrix(r,c);
-			end
-			% Fast way that doesn't actually work:
-			% jDoubleMatrix(r,:) = octaveMatrix(r,:);
-		end
+		jDoubleMatrix = octaveMatrix;
 	end
 
 end