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
 ******************************************************************************/

//
//  @author raver119@gmail.com
//

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

#include <ops/declarable/headers/shape.h>

namespace sd {
namespace ops {
CUSTOM_OP_IMPL(tile_to_shape, 1, 1, false, 0, -1) {
  auto input = INPUT_VARIABLE(0);
  auto output = OUTPUT_VARIABLE(0);

  std::vector<sd::LongType> outShape(block.getIArguments()->begin(), block.getIArguments()->end());

  if (block.isInplace()) {
    input->tileToShape(outShape, *input);
  } else {
    input->tileToShape(outShape, *output);
  }

  return sd::Status::OK;
}

DECLARE_SHAPE_FN(tile_to_shape) {
  auto in = inputShape->at(0);

  // output shape always equals to arguments

  auto conv = ArrayUtils::toLongVector(*block.getIArguments());

  auto newShape =
      ConstantShapeHelper::getInstance().createShapeInfo(ArrayOptions::dataType(in), shape::order(in), conv);

  return SHAPELIST(newShape);
}

DECLARE_TYPES(tile_to_shape) { getOpDescriptor()->setAllowedInputTypes(sd::DataType::ANY)->setSameMode(true); }

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

CUSTOM_OP_IMPL(tile_to_shape_bp, 2, 1, true, 0, -1) {
  auto input = INPUT_VARIABLE(0);
  auto epsNext = INPUT_VARIABLE(1);

  auto gradX = OUTPUT_VARIABLE(0);

  auto axisX = ShapeUtils::evalBroadcastBackwardAxis(input->shapeInfo(), epsNext->shapeInfo());
  // FIX ME: reduceAlongDimension should have a signature with result pass to avoid assigning twice
  if (!axisX.empty()) {
    auto tempRes = epsNext->reduceAlongDimension(reduce::Sum, &axisX);
    gradX->assign(tempRes);
  } else
    gradX->assign(epsNext);

  STORE_RESULT(gradX);

  return sd::Status::OK;
}

DECLARE_SHAPE_FN(tile_to_shape_bp) {
  auto in = inputShape->at(0);

  sd::LongType *newShape;
  COPY_SHAPE(in, newShape);

  return SHAPELIST(CONSTANT(newShape));
}
}  // namespace ops
}  // namespace sd

#endif