try to generalize more the transformations

src/main/java/io/bioimage/modelrunner/transformations/AbstractTensorPixelTransformation.java

@@ -20,10 +20,18 @@
 package io.bioimage.modelrunner.transformations;
 
 import io.bioimage.modelrunner.tensor.Tensor;
-
+import net.imglib2.RandomAccessibleInterval;
 import net.imglib2.loops.LoopBuilder;
 import net.imglib2.type.NativeType;
 import net.imglib2.type.numeric.RealType;
+import net.imglib2.type.numeric.integer.ByteType;
+import net.imglib2.type.numeric.integer.IntType;
+import net.imglib2.type.numeric.integer.LongType;
+import net.imglib2.type.numeric.integer.ShortType;
+import net.imglib2.type.numeric.integer.UnsignedByteType;
+import net.imglib2.type.numeric.integer.UnsignedIntType;
+import net.imglib2.type.numeric.integer.UnsignedShortType;
+import net.imglib2.type.numeric.real.DoubleType;
 import net.imglib2.type.numeric.real.FloatType;
 
 /**
@@ -38,6 +46,14 @@ public class AbstractTensorPixelTransformation extends AbstractTensorTransformat
 {
 
 	private FloatUnaryOperator fun;
+	private DoubleUnaryOperator dun;
+	private ByteUnaryOperator bun;
+	private UByteUnaryOperator ubun;
+	private ShortUnaryOperator sun;
+	private UShortUnaryOperator usun;
+	private IntUnaryOperator iun;
+	private UIntUnaryOperator uiun;
+	private LongUnaryOperator lun;
 
 	protected AbstractTensorPixelTransformation( final String name)
 	{
@@ -52,25 +68,116 @@ protected void setFloatUnitaryOperator(final FloatUnaryOperator fun) {
 	public < R extends RealType< R > & NativeType< R > > Tensor< FloatType > apply( final Tensor< R > input )
 	{
 		final Tensor< FloatType > output = makeOutput( input );
-		LoopBuilder
-				.setImages( input.getData(), output.getData() )
-				.multiThreaded()
-				.forEachPixel( ( i, o ) -> o.set( fun.applyAsFloat( i.getRealFloat() ) ) );
+		applyInPlace(output);
 		return output;
 	}
 
+	@SuppressWarnings("unchecked")
 	@Override
-	public void applyInPlace( final Tensor< FloatType > input )
+	public < R extends RealType< R > & NativeType< R > >
+	void applyInPlace( final Tensor< R > input )
 	{
-		LoopBuilder
-				.setImages( input.getData() )
-				.multiThreaded()
-				.forEachPixel( i -> i.set( fun.applyAsFloat( i.get() ) ) );
+		if (input.getData().getAt(0) instanceof FloatType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<FloatType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( fun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof DoubleType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<DoubleType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( dun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof ByteType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<ByteType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( bun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof UnsignedByteType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<UnsignedByteType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( ubun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof ShortType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<ShortType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( sun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof UnsignedShortType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<UnsignedShortType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( usun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof IntType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<IntType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( iun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof UnsignedIntType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<UnsignedIntType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( uiun.applyAs( i.get() ) ) );
+		} else if (input.getData().getAt(0) instanceof LongType) {
+			LoopBuilder
+			.setImages( (RandomAccessibleInterval<LongType>) input.getData() )
+			.multiThreaded()
+			.forEachPixel( i -> i.set( lun.applyAs( i.get() ) ) );
+		} else {
+			throw new IllegalArgumentException("Unsupported data type.");
+		}
 	}
 
 	@FunctionalInterface
 	public interface FloatUnaryOperator
 	{
-		float applyAsFloat( float in );
+		float applyAs( float in );
+	}
+
+	@FunctionalInterface
+	public interface DoubleUnaryOperator
+	{
+		float applyAs( double in );
+	}
+
+	@FunctionalInterface
+	public interface ByteUnaryOperator
+	{
+		byte applyAs( byte in );
+	}
+
+	@FunctionalInterface
+	public interface UByteUnaryOperator
+	{
+		int applyAs( int i );
+	}
+
+	@FunctionalInterface
+	public interface ShortUnaryOperator
+	{
+		short applyAs( short in );
+	}
+
+	@FunctionalInterface
+	public interface UShortUnaryOperator
+	{
+		int applyAs( int i );
+	}
+
+	@FunctionalInterface
+	public interface IntUnaryOperator
+	{
+		int applyAs( int in );
+	}
+
+	@FunctionalInterface
+	public interface UIntUnaryOperator
+	{
+		long applyAs( long in );
+	}
+
+	@FunctionalInterface
+	public interface LongUnaryOperator
+	{
+		long applyAs( long in );
 	}
 }

src/main/java/io/bioimage/modelrunner/transformations/AbstractTensorTransformation.java

@@ -36,7 +36,7 @@ public abstract class AbstractTensorTransformation implements TensorTransformati
 
 	private Mode mode = Mode.FIXED;
 
-	protected static String DEFAULT_MISSING_ARG_ERR = "Cannot execute Clip BioImage.io transformation because '%s' "
+	protected static String DEFAULT_MISSING_ARG_ERR = "Cannot execute %s BioImage.io transformation because '%s' "
 			+ "parameter was not set.";
 
 	protected AbstractTensorTransformation( final String name )