deeplearning4j/deeplearning4j

/* ******************************************************************************
 *
 *
 * This program and the accompanying materials are made available under the
 * terms of the Apache License, Version 2.0 which is available at
 * https://www.apache.org/licenses/LICENSE-2.0.
 *
 *  See the NOTICE file distributed with this work for additional
 *  information regarding copyright ownership.
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
 * License for the specific language governing permissions and limitations
 * under the License.
 *
 * SPDX-License-Identifier: Apache-2.0
 ******************************************************************************/

//
// Created by raver119 on 23.11.17.
//

#include <system/op_boilerplate.h>
#if NOT_EXCLUDED(OP_squaredsubtract)

#include <ops/declarable/CustomOperations.h>
#include <ops/declarable/generic/helpers/BroadcastHelper.h>

namespace sd {
namespace ops {
BROADCASTABLE_OP_IMPL(squaredsubtract, 0, 0) {
  auto x = INPUT_VARIABLE(0);
  auto y = INPUT_VARIABLE(1);
  auto z = OUTPUT_VARIABLE(0);

  BROADCAST_CHECK_EMPTY(x, y, z);

  auto tZ = BroadcastHelper::broadcastApply(BROADCAST(SquaredSubtract), x, y, z);
  if (tZ == nullptr)
    return sd::Status::KERNEL_FAILURE;
  else if (tZ != z) {
    OVERWRITE_RESULT(tZ);
  }

  return sd::Status::OK;
}
DECLARE_SYN(squareddifference, squaredsubtract);

DECLARE_TYPES(squaredsubtract) {
  getOpDescriptor()
      ->setAllowedInputTypes(0, DataType::ANY)
      ->setAllowedInputTypes(1, DataType::ANY)
      ->setAllowedOutputTypes(0, DataType::INHERIT);
}

CUSTOM_OP_IMPL(squaredsubtract_bp, 3, 2, false, 0, 0) {
  auto x = INPUT_VARIABLE(0);
  auto y = INPUT_VARIABLE(1);
  auto epsNext = INPUT_VARIABLE(2);

  auto gradX = OUTPUT_VARIABLE(0);
  auto gradY = OUTPUT_VARIABLE(1);


  auto ts = NDArrayFactory::create(x->dataType(), 2, block.launchContext());

  if (x->isSameShape(y)) {
    // PWT case case

    // X gradient
    gradX->assign((*epsNext) * ts * ((*x) - (*y)));

    // Y gradient
    gradY->assign((*epsNext) * ts * ((*y) - (*x)));

  } else if (y->isScalar()) {
    // scalar case
    auto tmpX = x->reduceNumber(reduce::Sum);
    gradY->assign(tmpX);
    gradX->assign((*epsNext) * ts * ((*x) - (*y)));
  } else {
    // broadcast case

    auto preX = x->dup();
    auto preY = y->dup();

    auto targetShape = epsNext->getShapeAsVector();

    preX.tileToShape(targetShape, preX);
    preY.tileToShape(targetShape, preY);

    auto resX = (*epsNext) * ts * ((*x) - (*y));
    preX.assign(resX);
    auto resY = (*epsNext) * ts * ((*y) - (*x));
    preY.assign(resY);

    auto axisX = ShapeUtils::evalBroadcastBackwardAxis(x->shapeInfo(), epsNext->shapeInfo());
    auto axisY = ShapeUtils::evalBroadcastBackwardAxis(y->shapeInfo(), epsNext->shapeInfo());

    if (axisX.size() > 0) {
      auto sum = preX.reduceAlongDimension(reduce::Sum, &axisX);
      gradX->assign(sum);
    } else
      gradX->assign(preX);

    if (axisY.size() > 0) {
      auto sum = preY.reduceAlongDimension(reduce::Sum, &axisY);
      gradY->assign(sum);
    } else
      gradY->assign(preY);
  }

  return sd::Status::OK;
}

DECLARE_SHAPE_FN(squaredsubtract_bp) {
  auto x = inputShape->at(0);
  auto y = inputShape->at(1);
  auto e = inputShape->at(2);

  // eps always has shape of x
  // grad always has shape of y

  sd::LongType *shapeE;
  sd::LongType *shapeG;

  COPY_SHAPE(x, shapeE);
  COPY_SHAPE(y, shapeG);

  return SHAPELIST(CONSTANT(shapeE), CONSTANT(shapeG));
}

DECLARE_TYPES(squaredsubtract_bp) {
  getOpDescriptor()->setAllowedInputTypes(DataType::ANY)->setAllowedOutputTypes({ALL_FLOATS});
}

}  // namespace ops
}  // namespace sd

#endif