Brutex
/
cavis
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

remove lengthLong (#236) 

Signed-off-by: Robert Altena <Rob@Ra-ai.com>
master
Robert Altena 2019-09-05 10:19:39 +09:00 committed by Alex Black
parent 03c52ef9dd
commit f25e3e71e5
13 changed files with 57 additions and 83 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/optimize/solvers/accumulation/EncodingHandler.java
View File

@ -135,7 +135,7 @@ public class EncodingHandler implements MessageHandler {
iterations.get().incrementAndGet();
if (boundary != null && atomicBoundary.get() < 0)
atomicBoundary.compareAndSet(-1, (int) (updates.lengthLong() * boundary));
atomicBoundary.compareAndSet(-1, (int) (updates.length() * boundary));
INDArray encoded;
@ -160,11 +160,11 @@ public class EncodingHandler implements MessageHandler {
double encLen = encoded.data().getInt(0);
// if updates are too dense - we fallback to bitmap encoding
if (encLen >= (updates.lengthLong() / 16)) {
if (encLen >= (updates.length() / 16)) {
log.debug("Switching back to bitmapEncoding: iteration {}, epoch {}, threshold {}, encoded length {}", iteration, epoch, currThreshold, encLen);
bitmapMode.get().set(true);
DataBuffer buffer = Nd4j.getDataBufferFactory().createInt(updates.lengthLong() / 16 + 5);
DataBuffer buffer = Nd4j.getDataBufferFactory().createInt(updates.length() / 16 + 5);
encoded = Nd4j.createArrayFromShapeBuffer(buffer, updates.shapeInfoDataBuffer());
Nd4j.getExecutioner().bitmapEncode(updates, encoded, currentThreshold.get().get());
@ -186,12 +186,12 @@ public class EncodingHandler implements MessageHandler {
}
} else {
//Dense bitmap updates
DataBuffer buffer = Nd4j.getDataBufferFactory().createInt(updates.lengthLong() / 16 + 5);
DataBuffer buffer = Nd4j.getDataBufferFactory().createInt(updates.length() / 16 + 5);
encoded = Nd4j.createArrayFromShapeBuffer(buffer, updates.shapeInfoDataBuffer());
long values = Nd4j.getExecutioner().bitmapEncode(updates, encoded, currentThreshold.get().get());
if (values < (updates.lengthLong() / 16 + 5) / 2) {
if (values < (updates.length() / 16 + 5) / 2) {
boolean current = bitmapMode.get().get();
bitmapMode.get().set(false);
if(!current) {

11
nd4j/nd4j-backends/nd4j-api-parent/nd4j-api/src/main/java/org/nd4j/linalg/api/ndarray/BaseNDArray.java
View File

@ -4641,17 +4641,6 @@ public abstract class BaseNDArray implements INDArray, Iterable {
return jvmShapeInfo.length;
}
/**
* Returns the total number of elements in the ndarray
*
* @return the number of elements in the ndarray
*/
@Override
@Deprecated
public long lengthLong() {
return jvmShapeInfo.length;
}
@Override
public INDArray broadcast(INDArray result) {
Nd4j.getCompressor().autoDecompress(this);

5
nd4j/nd4j-backends/nd4j-api-parent/nd4j-api/src/main/java/org/nd4j/linalg/api/ndarray/BaseSparseNDArray.java
View File

@ -279,11 +279,6 @@ public abstract class BaseSparseNDArray implements ISparseNDArray {
return (int) length();
}
@Override
public long lengthLong() {
return length;
}
protected void init(long[] shape) {
if (shape.length == 1) {

12
nd4j/nd4j-backends/nd4j-api-parent/nd4j-api/src/main/java/org/nd4j/linalg/api/ndarray/INDArray.java
View File

@ -2377,17 +2377,7 @@ public interface INDArray extends Serializable, AutoCloseable {
* @return the number of elements in the ndarray
*/
long length();
/**
* Returns the total number of elements in the ndarray
*
* @return the number of elements in the ndarray
* @deprecated use {@link #length()}
*/
@Deprecated
long lengthLong();
/**
* Broadcasts this ndarray to be the specified shape
*

12
nd4j/nd4j-backends/nd4j-backend-impls/nd4j-cuda/src/main/java/org/nd4j/linalg/jcublas/JCublasNDArray.java
View File

@ -664,7 +664,7 @@ public class JCublasNDArray extends BaseNDArray {
//if (1 < 0) {
Nd4j.getExecutioner().commit();
DataBuffer buffer = Nd4j.createBuffer(this.lengthLong(), false);
DataBuffer buffer = Nd4j.createBuffer(this.length(), false);
AllocationPoint pointDst = AtomicAllocator.getInstance().getAllocationPoint(buffer);
AllocationPoint pointSrc = AtomicAllocator.getInstance().getAllocationPoint(this.data);
@ -686,10 +686,10 @@ public class JCublasNDArray extends BaseNDArray {
val perfD = PerformanceTracker.getInstance().helperStartTransaction();
if (pointSrc.isActualOnDeviceSide()) {
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getDevicePointer(), this.lengthLong() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyDeviceToDevice, context.getOldStream()) == 0)
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getDevicePointer(), this.length() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyDeviceToDevice, context.getOldStream()) == 0)
throw new ND4JIllegalStateException("memcpyAsync failed");
} else {
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getHostPointer(), this.lengthLong() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyHostToDevice, context.getOldStream()) == 0)
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getHostPointer(), this.length() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyHostToDevice, context.getOldStream()) == 0)
throw new ND4JIllegalStateException("memcpyAsync failed");
direction = MemcpyDirection.HOST_TO_DEVICE;
@ -738,7 +738,7 @@ public class JCublasNDArray extends BaseNDArray {
if (!this.isView()) {
Nd4j.getExecutioner().commit();
val buffer = Nd4j.createBuffer(this.dataType(), this.lengthLong(), false);
val buffer = Nd4j.createBuffer(this.dataType(), this.length(), false);
val pointDst = AtomicAllocator.getInstance().getAllocationPoint(buffer);
val pointSrc = AtomicAllocator.getInstance().getAllocationPoint(this.data);
@ -749,10 +749,10 @@ public class JCublasNDArray extends BaseNDArray {
val perfD = PerformanceTracker.getInstance().helperStartTransaction();
if (pointSrc.isActualOnDeviceSide()) {
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getDevicePointer(), this.lengthLong() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyDeviceToDevice, context.getOldStream()) == 0)
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getDevicePointer(), this.length() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyDeviceToDevice, context.getOldStream()) == 0)
throw new ND4JIllegalStateException("memcpyAsync failed");
} else {
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getHostPointer(), this.lengthLong() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyHostToDevice, context.getOldStream()) == 0)
if (NativeOpsHolder.getInstance().getDeviceNativeOps().memcpyAsync(pointDst.getDevicePointer(), pointSrc.getHostPointer(), this.length() * Nd4j.sizeOfDataType(buffer.dataType()), CudaConstants.cudaMemcpyHostToDevice, context.getOldStream()) == 0)
throw new ND4JIllegalStateException("memcpyAsync failed");
direction = MemcpyDirection.HOST_TO_DEVICE;

22
nd4j/nd4j-backends/nd4j-backend-impls/nd4j-cuda/src/main/java/org/nd4j/linalg/jcublas/JCublasNDArrayFactory.java
View File

@ -424,7 +424,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
val perfD = PerformanceTracker.getInstance().helperStartTransaction();
nativeOps.memcpyAsync(point.getDevicePointer(), point.getHostPointer(), ret.lengthLong() * Nd4j.sizeOfDataType(ret.data().dataType()), CudaConstants.cudaMemcpyHostToDevice, context.getSpecialStream());
nativeOps.memcpyAsync(point.getDevicePointer(), point.getHostPointer(), ret.length() * Nd4j.sizeOfDataType(ret.data().dataType()), CudaConstants.cudaMemcpyHostToDevice, context.getSpecialStream());
context.getSpecialStream().synchronize();
if (nativeOps.lastErrorCode() != 0)
@ -580,7 +580,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
if (true) {
Nd4j.getExecutioner().push();
long len = target.lengthLong();
long len = target.length();
AtomicAllocator allocator = AtomicAllocator.getInstance();
@ -598,7 +598,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays[i].elementWiseStride() != 1)
throw new ND4JIllegalStateException("Native averaging is applicable only to continuous INDArrays");
if (arrays[i].lengthLong() != len)
if (arrays[i].length() != len)
throw new ND4JIllegalStateException("All arrays should have equal length for averaging");
AllocationPoint point = allocator.getAllocationPoint(arrays[i]);
@ -621,7 +621,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
return target;
} else {
long len = target.lengthLong();
long len = target.length();
Nd4j.getExecutioner().commit();
@ -637,7 +637,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays[i].elementWiseStride() != 1)
throw new ND4JIllegalStateException("Native averaging is applicable only to continuous INDArrays");
if (arrays[i].lengthLong() != len)
if (arrays[i].length() != len)
throw new ND4JIllegalStateException("All arrays should have equal length for averaging");
((BaseCudaDataBuffer) arrays[i].data()).lazyAllocateHostPointer();
@ -689,7 +689,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
Nd4j.getExecutioner().push();
long len = target != null ? target.lengthLong() : arrays[0].lengthLong();
long len = target != null ? target.length() : arrays[0].length();
AtomicAllocator allocator = AtomicAllocator.getInstance();
@ -707,7 +707,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays[i].elementWiseStride() != 1)
throw new ND4JIllegalStateException("Native averaging is applicable only to continuous INDArrays");
if (arrays[i].lengthLong() != len)
if (arrays[i].length() != len)
throw new ND4JIllegalStateException("All arrays should have equal length for averaging");
AllocationPoint point = allocator.getAllocationPoint(arrays[i]);
@ -744,7 +744,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
/**
* We expect all operations are complete at this point
*/
long len = target == null ? arrays[0].lengthLong() : target.lengthLong();
long len = target == null ? arrays[0].length() : target.length();
val context = (CudaContext) AtomicAllocator.getInstance().getDeviceContext();
@ -758,7 +758,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays[i].elementWiseStride() != 1)
throw new ND4JIllegalStateException("Native averaging is applicable only to continuous INDArrays");
if (arrays[i].lengthLong() != len)
if (arrays[i].length() != len)
throw new ND4JIllegalStateException("All arrays should have equal length for averaging");
((BaseCudaDataBuffer) arrays[i].data()).lazyAllocateHostPointer();
@ -1303,7 +1303,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
}
int numTads = (int)(tensor.lengthLong() / tadLength);
int numTads = (int)(tensor.length() / tadLength);
INDArray[] result = new INDArray[numTads];
long[] xPointers = new long[numTads];
@ -1378,7 +1378,7 @@ public class JCublasNDArrayFactory extends BaseNativeNDArrayFactory {
new CudaPointer(0));
// we're sending > 10m elements to radixSort
boolean isRadix = !x.isView() && (x.lengthLong() > 1024 * 1024 * 10);
boolean isRadix = !x.isView() && (x.length() > 1024 * 1024 * 10);
INDArray tmpX = x;
// we need to guarantee all threads are finished here

26
nd4j/nd4j-backends/nd4j-backend-impls/nd4j-cuda/src/main/java/org/nd4j/linalg/jcublas/ops/executioner/CudaExecutioner.java
View File

@ -293,9 +293,9 @@ public class CudaExecutioner extends DefaultOpExecutioner {
Pointer yDevTadShapeInfo = null;
if (op.y() != null) {
if (dimension.length == 0 || (dimension.length == 1 && dimension[0] == Integer.MAX_VALUE )|| op.x().tensorAlongDimension(0, dimension).lengthLong() != op.y().lengthLong()) {
if (!op.isComplexAccumulation() && op.x().lengthLong() != op.y().lengthLong())
throw new ND4JIllegalStateException("Op.X [" + op.x().lengthLong() + "] and Op.Y [" + op.y().lengthLong() + "] lengths should match");
if (dimension.length == 0 || (dimension.length == 1 && dimension[0] == Integer.MAX_VALUE )|| op.x().tensorAlongDimension(0, dimension).length() != op.y().length()) {
if (!op.isComplexAccumulation() && op.x().length() != op.y().length())
throw new ND4JIllegalStateException("Op.X [" + op.x().length() + "] and Op.Y [" + op.y().length() + "] lengths should match");
if (!op.z().isScalar()) {
Pair<DataBuffer, DataBuffer> yTadBuffers = tadManager.getTADOnlyShapeInfo(op.y(), dimension);
@ -536,7 +536,7 @@ public class CudaExecutioner extends DefaultOpExecutioner {
} else {
if (op.y() != null) {
//2 options here: either pairwise, equal sizes - OR every X TAD vs. entirety of Y
if (op.x().lengthLong() == op.y().lengthLong()) {
if (op.x().length() == op.y().length()) {
//Pairwise
if (!wholeDims && op.x().tensorsAlongDimension(dimension) != op.y().tensorsAlongDimension(dimension)) {
throw new ND4JIllegalStateException("Number of TADs along dimension don't match: (x shape = " +
@ -548,11 +548,11 @@ public class CudaExecutioner extends DefaultOpExecutioner {
throw new ND4JIllegalStateException("TAD vs TAD comparison requires dimension (or other comparison mode was supposed to be used?)");
//Every X TAD vs. entirety of Y
val xTADSize = op.x().lengthLong() / op.x().tensorsAlongDimension(dimension);
val xTADSize = op.x().length() / op.x().tensorsAlongDimension(dimension);
if (xTADSize != op.y().length()) {
throw new ND4JIllegalStateException("Size of TADs along dimension don't match for pairwise execution:" +
" (x TAD size = " + xTADSize + ", y size = " + op.y().lengthLong());
" (x TAD size = " + xTADSize + ", y size = " + op.y().length());
}
}
}
@ -976,7 +976,7 @@ public class CudaExecutioner extends DefaultOpExecutioner {
if (op.y() != null) {
//2 options here: either pairwise, equal sizes - OR every X TAD vs. entirety of Y
if (op.x().lengthLong() == op.y().lengthLong()) {
if (op.x().length() == op.y().length()) {
//Pairwise
if (op.x().tensorsAlongDimension(dimension) != op.y().tensorsAlongDimension(dimension)) {
throw new ND4JIllegalStateException("Number of TADs along dimension don't match: (x shape = " +
@ -985,11 +985,11 @@ public class CudaExecutioner extends DefaultOpExecutioner {
}
} else {
//Every X TAD vs. entirety of Y
val xTADSize = op.x().lengthLong() / op.x().tensorsAlongDimension(dimension);
val xTADSize = op.x().length() / op.x().tensorsAlongDimension(dimension);
if (xTADSize != op.y().length()) {
throw new ND4JIllegalStateException("Size of TADs along dimension don't match for pairwise execution:" +
" (x TAD size = " + xTADSize + ", y size = " + op.y().lengthLong());
" (x TAD size = " + xTADSize + ", y size = " + op.y().length());
}
}
}
@ -2031,8 +2031,8 @@ public class CudaExecutioner extends DefaultOpExecutioner {
long compressedLength = buffer.getInt(0);
long originalLength = buffer.getInt(1);
if (target.lengthLong() != originalLength)
throw new ND4JIllegalStateException("originalLength ["+ originalLength+"] stored in encoded array doesn't match target length ["+ target.lengthLong()+"]");
if (target.length() != originalLength)
throw new ND4JIllegalStateException("originalLength ["+ originalLength+"] stored in encoded array doesn't match target length ["+ target.length()+"]");
DataBuffer result = target.data();
@ -2056,7 +2056,7 @@ public class CudaExecutioner extends DefaultOpExecutioner {
@Override
public long bitmapEncode(INDArray indArray, INDArray target, double threshold) {
long length = indArray.lengthLong();
long length = indArray.length();
long tLen = target.data().length();
if (tLen != (length / 16 + 5))
@ -2117,7 +2117,7 @@ public class CudaExecutioner extends DefaultOpExecutioner {
context.getBufferScalar(),
context.getBufferReduction());
nativeOps.decodeBitmap(extras, AtomicAllocator.getInstance().getPointer(encoded.data(), context), target.lengthLong(), AtomicAllocator.getInstance().getPointer(target, context), (LongPointer) AtomicAllocator.getInstance().getHostPointer(target.shapeInfoDataBuffer()));
nativeOps.decodeBitmap(extras, AtomicAllocator.getInstance().getPointer(encoded.data(), context), target.length(), AtomicAllocator.getInstance().getPointer(target, context), (LongPointer) AtomicAllocator.getInstance().getHostPointer(target.shapeInfoDataBuffer()));
if (nativeOps.lastErrorCode() != 0)
throw new RuntimeException(nativeOps.lastErrorMessage());

2
nd4j/nd4j-backends/nd4j-backend-impls/nd4j-cuda/src/main/java/org/nd4j/linalg/jcublas/ops/executioner/CudaGridExecutioner.java
View File

@ -655,7 +655,7 @@ public class CudaGridExecutioner extends CudaExecutioner implements GridExecutio
op.setZ(ret);
} else {
// compare length
if (op.z().lengthLong() != ArrayUtil.prodLong(retShape))
if (op.z().length() != ArrayUtil.prodLong(retShape))
throw new ND4JIllegalStateException("Shape of target array for reduction [" + Arrays.toString(op.z().shape()) + "] doesn't match expected [" + Arrays.toString(retShape) + "]");
ret = op.z();

10
nd4j/nd4j-backends/nd4j-backend-impls/nd4j-native/src/main/java/org/nd4j/linalg/cpu/nativecpu/CpuNDArrayFactory.java
View File

@ -514,7 +514,7 @@ public class CpuNDArrayFactory extends BaseNativeNDArrayFactory {
int numTads = (int)(tensor.lengthLong() / tadLength);
int numTads = (int)(tensor.length() / tadLength);
INDArray[] result = new INDArray[numTads];
PointerPointer targets = new PointerPointer(numTads);
@ -693,7 +693,7 @@ public class CpuNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays.length == 1)
return target.addi(arrays[0]);
long len = target.lengthLong();
long len = target.length();
PointerPointer dataPointers = new PointerPointer(arrays.length);
@ -703,7 +703,7 @@ public class CpuNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays[i].elementWiseStride() != 1)
throw new ND4JIllegalStateException("Native accumulation is applicable only to continuous INDArrays");
if (arrays[i].lengthLong() != len)
if (arrays[i].length() != len)
throw new ND4JIllegalStateException("All arrays should have equal length for accumulation");
dataPointers.put(i, arrays[i].data().addressPointer());
@ -744,7 +744,7 @@ public class CpuNDArrayFactory extends BaseNativeNDArrayFactory {
return target.assign(arrays[0]);
}
long len = target != null ? target.lengthLong() : arrays[0].length();
long len = target != null ? target.length() : arrays[0].length();
PointerPointer dataPointers = new PointerPointer(arrays.length);
val firstType = arrays[0].dataType();
@ -757,7 +757,7 @@ public class CpuNDArrayFactory extends BaseNativeNDArrayFactory {
if (arrays[i].elementWiseStride() != 1)
throw new ND4JIllegalStateException("Native averaging is applicable only to continuous INDArrays");
if (arrays[i].lengthLong() != len)
if (arrays[i].length() != len)
throw new ND4JIllegalStateException("All arrays should have equal length for averaging");
dataPointers.put(i, arrays[i].data().addressPointer());

22
nd4j/nd4j-backends/nd4j-backend-impls/nd4j-native/src/main/java/org/nd4j/linalg/cpu/nativecpu/ops/NativeOpExecutioner.java
View File

@ -303,11 +303,11 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
}
} else {
//Every X TAD vs. entirety of Y
val xTADSize = op.x().lengthLong() / op.x().tensorsAlongDimension(dimension);
val xTADSize = op.x().length() / op.x().tensorsAlongDimension(dimension);
if (xTADSize != op.y().length()) {
throw new ND4JIllegalStateException("Size of TADs along dimension don't match for pairwise execution:" +
" (x TAD size = " + xTADSize + ", y size = " + op.y().lengthLong());
" (x TAD size = " + xTADSize + ", y size = " + op.y().length());
}
}
}
@ -329,7 +329,7 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
long xT = op.x().tensorsAlongDimension(dimension);
long yT = op.y().tensorsAlongDimension(dimension);
if (op.z().lengthLong() != xT * yT)
if (op.z().length() != xT * yT)
throw new ND4JIllegalStateException("Shape of target array for reduction [" + Arrays.toString(op.z().shape()) + "] doesn't match expected [" + (xT * yT) + "]");
}
@ -358,7 +358,7 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
// we're going to check, if that's TAD vs TAD comparison or TAD vs full array. if later - we're going slightly different route
boolean tvf = false;
if (op.y() != null) {
if (op.x().tensorAlongDimension(0, dimension).lengthLong() == op.y().lengthLong()) {
if (op.x().tensorAlongDimension(0, dimension).length() == op.y().length()) {
tvf = true;
}
}
@ -366,10 +366,10 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
if (op.isComplexAccumulation()) {
yTadBuffers = tadManager.getTADOnlyShapeInfo(op.y(), dimension);
if (op.x().tensorAlongDimension(0, dimension).lengthLong() != op.y().tensorAlongDimension(0, dimension).lengthLong())
if (op.x().tensorAlongDimension(0, dimension).length() != op.y().tensorAlongDimension(0, dimension).length())
throw new ND4JIllegalStateException("Impossible to issue AllDistances operation: TAD lengths mismatch along given dimension: " +
"x TAD length = " + op.x().tensorAlongDimension(0, dimension).lengthLong() + ", y TAD length " +
op.y().tensorAlongDimension(0, dimension).lengthLong());
"x TAD length = " + op.x().tensorAlongDimension(0, dimension).length() + ", y TAD length " +
op.y().tensorAlongDimension(0, dimension).length());
}
/**
@ -659,7 +659,7 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
//validateDataType(Nd4j.dataType(), op);
if (op.x().lengthLong() != op.z().lengthLong())
if (op.x().length() != op.z().length())
throw new ND4JIllegalStateException("op.X length should be equal to op.Z length: " +
"x.length()=" + op.x().length() + ", z.length()=" + op.z().length() + " - x shape info = ["
+ Arrays.toString(op.x().shapeInfoDataBuffer().asInt()) + "], z shape info = ["
@ -1449,8 +1449,8 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
long originalLength = buffer.getInt(1);
float threshold = buffer.getInt(2);
if (target.lengthLong() != originalLength)
throw new ND4JIllegalStateException("originalLength ["+ originalLength+"] stored in encoded array doesn't match target length ["+ target.lengthLong()+"]");
if (target.length() != originalLength)
throw new ND4JIllegalStateException("originalLength ["+ originalLength+"] stored in encoded array doesn't match target length ["+ target.length()+"]");
DataTypeEx typeDst = AbstractCompressor.getBufferTypeEx(target.data());
@ -1465,7 +1465,7 @@ public class NativeOpExecutioner extends DefaultOpExecutioner {
@Override
public long bitmapEncode(INDArray indArray, INDArray target, double threshold) {
long length = indArray.lengthLong();
long length = indArray.length();
long tLen = target.data().length();
if (tLen != (length / 16 + 5))

2
nd4j/nd4j-backends/nd4j-tests/src/test/java/org/nd4j/linalg/Nd4jTestsC.java
View File

@ -5155,7 +5155,7 @@ public class Nd4jTestsC extends BaseNd4jTest {
INDArray res = x.entropy(1);
assertEquals(10, res.lengthLong());
assertEquals(10, res.length());
for (int t = 0; t < x.rows(); t++) {
double exp = MathUtils.entropy(x.getRow(t).dup().data().asDouble());

4
nd4j/nd4j-backends/nd4j-tests/src/test/java/org/nd4j/linalg/ShufflesTests.java
View File

@ -415,7 +415,7 @@ public class ShufflesTests extends BaseNd4jTest {
for (int x = 0; x < newData.rows(); x++) {
INDArray row = newData.getRow(x);
for (int y = 0; y < row.lengthLong(); y++) {
for (int y = 0; y < row.length(); y++) {
if (Math.abs(row.getFloat(y) - newMap[x]) > Nd4j.EPS_THRESHOLD) {
System.out.print("Different data in a row");
return false;
@ -442,7 +442,7 @@ public class ShufflesTests extends BaseNd4jTest {
for (int x = 0; x < newData.rows(); x++) {
INDArray column = newData.getColumn(x);
double val = column.getDouble(0);
for (int y = 0; y < column.lengthLong(); y++) {
for (int y = 0; y < column.length(); y++) {
if (Math.abs(column.getFloat(y) - val) > Nd4j.EPS_THRESHOLD) {
System.out.print("Different data in a column: " + column.getFloat(y));
return false;

2
nd4j/nd4j-backends/nd4j-tests/src/test/java/org/nd4j/linalg/compression/CompressionTests.java
View File

@ -168,7 +168,7 @@ public class CompressionTests extends BaseNd4jTest {
INDArray decompressed = Nd4j.create(1, initial.length());
Nd4j.getExecutioner().thresholdDecode(compressed, decompressed);
log.info("Decompressed length: {}", decompressed.lengthLong());
log.info("Decompressed length: {}", decompressed.length());
assertEquals(exp_d, decompressed);
}