org.nd4j.linalg.factory.Nd4j.trueScalar()方法的使用及代码示例

x33g5p2x  于2022-01-24 转载在 其他  
字(7.5k)|赞(0)|评价(0)|浏览(129)

本文整理了Java中org.nd4j.linalg.factory.Nd4j.trueScalar()方法的一些代码示例，展示了Nd4j.trueScalar()的具体用法。这些代码示例主要来源于Github/Stackoverflow/Maven等平台，是从一些精选项目中提取出来的代码，具有较强的参考意义，能在一定程度帮忙到你。Nd4j.trueScalar()方法的具体详情如下：
包路径：org.nd4j.linalg.factory.Nd4j
类名称：Nd4j
方法名：trueScalar

Nd4j.trueScalar介绍

[英]This method creates new 0D INDArray, aka scalar. PLEASE NOTE: Temporary method, added to ensure backward compatibility
[中]

代码示例

代码示例来源：origin: deeplearning4j/nd4j

  1. @Override
  2. public INDArray getScalar(long... indexes) {
  3.   return Nd4j.trueScalar(getDouble(indexes));
  4. }

代码示例来源：origin: deeplearning4j/nd4j

  1. public static INDArray createUninitializedDetached(long[] shape, char ordering) {
  2.   if (shape.length == 0)
  3.     return trueScalar(0.0);
  4.   //ensure shapes that wind up being scalar end up with the write shape
  5.   if (shape.length == 1 && shape[0] == 0) {
  6.     shape = new long[] {1, 1};
  7.   } else if (shape.length == 1) {
  8.     shape = new long[] {1, shape[0]};
  9.   }
  10.   checkShapeValues(shape);
  11.   INDArray ret = INSTANCE.createUninitializedDetached(shape, ordering);
  12.   logCreationIfNecessary(ret);
  13.   return ret;
  14. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Creates an ndarray with the specified value
  3.  * as the  only value in the ndarray.
  4.  * Some people may know this as np.full
  5.  *
  6.  * @param shape the shape of the ndarray
  7.  * @param value the value to assign
  8.  * @return the created ndarray
  9.  */
  10. public static INDArray valueArrayOf(int[] shape, double value) {
  11.   if (shape.length == 0)
  12.     return trueScalar(value);
  13.   checkShapeValues(shape);
  14.   INDArray ret = INSTANCE.valueArrayOf(shape, value);
  15.   logCreationIfNecessary(ret);
  16.   return ret;
  17. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Creates an ndarray with the specified value
  3.  * as the  only value in the ndarray.
  4.  * Some people may know this as np.full
  5.  *
  6.  * @param shape the shape of the ndarray
  7.  * @param value the value to assign
  8.  * @return the created ndarray
  9.  */
  10. public static INDArray valueArrayOf(long[] shape, double value) {
  11.   if (shape.length == 0)
  12.     return trueScalar(value);
  13.   checkShapeValues(shape);
  14.   INDArray ret = INSTANCE.valueArrayOf(shape, value);
  15.   logCreationIfNecessary(ret);
  16.   return ret;
  17. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Cretes uninitialized INDArray detached from any (if any) workspace
  3.  *
  4.  * @param shape
  5.  * @param ordering
  6.  * @return
  7.  */
  8. public static INDArray createUninitializedDetached(int[] shape, char ordering) {
  9.   if (shape.length == 0)
  10.     return trueScalar(0.0);
  11.   //ensure shapes that wind up being scalar end up with the write shape
  12.   if (shape.length == 1 && shape[0] == 0) {
  13.     shape = new int[] {1, 1};
  14.   } else if (shape.length == 1) {
  15.     shape = new int[] {1, shape[0]};
  16.   }
  17.   checkShapeValues(shape);
  18.   INDArray ret = INSTANCE.createUninitializedDetached(shape, ordering);
  19.   logCreationIfNecessary(ret);
  20.   return ret;
  21. }

代码示例来源：origin: deeplearning4j/nd4j

  1. String shapeString = line.split(":")[1].replace("[", "").replace("],", "");
  2. if (shapeString.isEmpty()) {
  3.   newArr = Nd4j.trueScalar(0);
  4. } else {
  5.   String[] shapeArr = shapeString.split(",");

代码示例来源：origin: deeplearning4j/nd4j

  1. public static INDArray createUninitialized(long[] shape, char ordering) {
  2.   if (shape.length == 0)
  3.     return trueScalar(0.0);
  4.   shape = getEnsuredShape(shape);
  5.   // now we allow 1D vectors
  6.   /*else if (shape.length == 1) {
  7.     shape = new int[] {1, shape[0]};
  8.   }
  9.   */
  10.   checkShapeValues(shape);
  11.   INDArray ret = INSTANCE.createUninitialized(shape, ordering);
  12.   logCreationIfNecessary(ret);
  13.   return ret;
  14. }

代码示例来源：origin: deeplearning4j/nd4j

  1. public static INDArray create(float[] data, long[] shape) {
  2.   if (shape.length == 0 && data.length == 1) {
  3.     return trueScalar(data[0]);
  4.   }
  5.   shape = getEnsuredShape(shape);
  6.   if (shape.length == 1) {
  7.     if (shape[0] != data.length)
  8.       throw new ND4JIllegalStateException("Shape of the new array doesn't match data length");
  9.   }
  10.   checkShapeValues(data.length, shape);
  11.   INDArray ret = INSTANCE.create(data, shape);
  12.   logCreationIfNecessary(ret);
  13.   return ret;
  14. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Creates an ndarray with the specified shape
  3.  *
  4.  * @param shape  the shape of the ndarray
  5.  * @param stride the stride for the ndarray
  6.  * @param offset the offset of the ndarray
  7.  * @return the instance
  8.  */
  9. public static INDArray create(double[] data, int[] shape, int[] stride, long offset, char ordering) {
  10.   if (data.length == 1 && shape.length == 0)
  11.     return trueScalar(data[0]);
  12.   shape = getEnsuredShape(shape);
  13.   if (shape.length == 1) {
  14.     if (shape[0] != data.length)
  15.       throw new ND4JIllegalStateException("Shape of the new array " + Arrays.toString(shape)
  16.           + " doesn't match data length: " + data.length);
  17.   }
  18.   checkShapeValues(data.length, shape);
  19.   INDArray ret = INSTANCE.create(data, shape, stride, offset, ordering);
  20.   logCreationIfNecessary(ret);
  21.   return ret;
  22. }

代码示例来源：origin: deeplearning4j/nd4j

  1. if (rank == 0) {
  2.   all.add(new Pair<>(Nd4j.trueScalar(Nd4j.rand(1, 1).getDouble(0)), "{}"));
  3.   return all;

代码示例来源：origin: deeplearning4j/nd4j

  1. public static INDArray create(double[] data, long[] shape) {
  2.   if (shape.length == 0 && data.length == 1) {
  3.     return trueScalar(data[0]);
  4.   }
  5.   shape = getEnsuredShape(shape);
  6.   if (shape.length == 1) {
  7.     if (shape[0] != data.length)
  8.       throw new ND4JIllegalStateException("Shape of the new array doesn't match data length");
  9.   }
  10.   checkShapeValues(data.length, shape);
  11.   INDArray ret = INSTANCE.create(data, shape);
  12.   logCreationIfNecessary(ret);
  13.   return ret;
  14. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Creates an *uninitialized* ndarray with the specified shape and ordering.<br>
  3.  * <b>NOTE</b>: The underlying memory (DataBuffer) will not be initialized. Don't use this unless you know what you are doing.
  4.  *
  5.  * @param shape the shape of the ndarray
  6.  * @param ordering the order of the ndarray
  7.  * @return the instance
  8.  */
  9. public static INDArray createUninitialized(int[] shape, char ordering) {
  10.   if (shape.length == 0)
  11.     return trueScalar(0.0);
  12.   shape = getEnsuredShape(shape);
  13.   // now we allow 1D vectors
  14.   /*else if (shape.length == 1) {
  15.     shape = new int[] {1, shape[0]};
  16.   }
  17.   */
  18.   checkShapeValues(shape);
  19.   INDArray ret = INSTANCE.createUninitialized(shape, ordering);
  20.   logCreationIfNecessary(ret);
  21.   return ret;
  22. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Create an ndrray with the specified shape
  3.  *
  4.  * @param data  the data to use with tne ndarray
  5.  * @param shape the shape of the ndarray
  6.  * @return the created ndarray
  7.  */
  8. public static INDArray create(float[] data, int[] shape) {
  9.   if (shape.length == 0 && data.length == 1) {
  10.     return trueScalar(data[0]);
  11.   }
  12.   shape = getEnsuredShape(shape);
  13.   if (shape.length == 1) {
  14.     if (shape[0] != data.length)
  15.       throw new ND4JIllegalStateException("Shape of the new array doesn't match data length");
  16.   }
  17.   checkShapeValues(data.length, shape);
  18.   INDArray ret = INSTANCE.create(data, shape);
  19.   logCreationIfNecessary(ret);
  20.   return ret;
  21. }

代码示例来源：origin: deeplearning4j/nd4j

  1. /**
  2.  * Note that iArgs (integer arguments) and  tArgs(double/float arguments)
  3.  * may end up being used under the following conditions:
  4.  * scalar operations (if a scalar is specified the you do not need to specify an ndarray)
  5.  * otherwise, if an ndarray is needed as a second input then put it in the inputs
  6.  *
  7.  * Usually, you only need 1 input (the equivalent of the array you're trying to do indexing on)
  8.  *
  9.  * @param inputs the inputs in to the op
  10.  * @param iArgs the integer arguments as needed
  11.  * @param tArgs the arguments
  12.  * @param condition the condition to filter on
  13.  */
  14. public Choose(INDArray[] inputs,List<Integer> iArgs, List<Double> tArgs,Condition condition) {
  15.   super(null, inputs, null);
  16.   if(condition == null) {
  17.     throw new ND4JIllegalArgumentException("Must specify a condition.");
  18.   }
  19.   if(!iArgs.isEmpty())
  20.     addIArgument(Ints.toArray(iArgs));
  21.   if(!tArgs.isEmpty())
  22.     addTArgument(Doubles.toArray(tArgs));
  23.   addIArgument(condition.condtionNum());
  24.   addOutputArgument(Nd4j.create(inputs[0].shape(), inputs[0].ordering()),Nd4j.trueScalar(1.0));
  25. }

代码示例来源：origin: deeplearning4j/nd4j

  1. return Nd4j.trueScalar(0.0);
  2.     return Nd4j.trueScalar(fa[0]);
  3.     return Nd4j.trueScalar(fa[0]);
  4.     return Nd4j.trueScalar(0.0);
  5.   INDArray array = Nd4j.trueScalar(val);
  6.   return array;
  7. } else if (tfTensor.getDoubleValCount() > 0) {
  8.     return Nd4j.trueScalar(da[0]);
  9.     return Nd4j.trueScalar(0.0);
  10.   INDArray array = Nd4j.trueScalar(val);
  11.   return array;
  12. } else if (tfTensor.getInt64ValCount() > 0)  {
  13.     return Nd4j.trueScalar(fa[0]);

代码示例来源：origin: improbable-research/keanu

  1. public static INDArray scalar(double scalarValue, DataBuffer.Type bufferType) {
  2.   Nd4j.setDataType(bufferType);
  3.   return Nd4j.trueScalar(scalarValue);
  4. }

内容来源于网络，如有侵权，请联系作者删除！

Nd4j

相关文章

    查看更多

    热门标签

    更多
    JavaquerypythonNode开发语言requestUtil数据库Table后端算法LoggerMessageElementParser

    最新文章

    更多

    Nd4j类方法

    • 蜀ICP备13028337号-1 大数据知识库 https://www.saoniuhuo.com © All rights reserved
    • 本站内容来源互联网,如果侵犯您的权益请联系我们删除, 联系方式：448109455@qq.com