Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance.
Project price only 1 $
You can buy this project and download/modify it how often you want.
org.nd4j.linalg.api.ops.impl.reduce.custom.BaseDynamicCustomReduction Maven / Gradle / Ivy
/*
* ******************************************************************************
* *
* *
* * 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
* *****************************************************************************
*/
package org.nd4j.linalg.api.ops.impl.reduce.custom;
import lombok.Getter;
import lombok.Setter;
import org.nd4j.autodiff.samediff.SDVariable;
import org.nd4j.autodiff.samediff.SameDiff;
import org.nd4j.common.base.Preconditions;
import org.nd4j.linalg.api.ndarray.INDArray;
import org.nd4j.linalg.api.ops.DynamicCustomOp;
import org.nd4j.linalg.api.ops.OpContext;
import java.util.Arrays;
import java.util.List;
/**
* Base class for reduction.
* 3 main properties matter for any sub class:
* 1. Dimensions can either be an input variable (SDVariable/INDArray) or int arguments.
* 2. If you want the dimensions passed as int arguments, pass in the int dimensions array as a constructor.
* 3. Keep dimensions preserves the rank of the output array relative to the input
* even when doing a reduce.
*
* @author Adam Gibson
*/
public abstract class BaseDynamicCustomReduction extends DynamicCustomOp {
@Setter
@Getter
protected boolean keepDims = false;
protected boolean isComplex = false;
@Setter @Getter
protected boolean isEmptyReduce = false;
protected int[] dimensions;
public BaseDynamicCustomReduction() {}
public BaseDynamicCustomReduction(SameDiff sameDiff,
SDVariable[] args,
boolean keepDims) {
this(sameDiff,args,keepDims,false);
}
public BaseDynamicCustomReduction(SameDiff sameDiff,
SDVariable[] args,
boolean keepDims,
int[] dimensions) {
this(sameDiff,args,keepDims,false,dimensions);
}
public BaseDynamicCustomReduction(SameDiff sameDiff,
SDVariable[] args,
boolean keepDims,
boolean isComplex) {
super(null,sameDiff,args);
this.isComplex = isComplex;
this.keepDims = keepDims;
addArgs();
}
public BaseDynamicCustomReduction(SameDiff sameDiff,
SDVariable[] args,
boolean keepDims,
boolean isComplex,
int[] dimensions) {
super(null,sameDiff,args);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(INDArray[] inputs, INDArray[] outputs) {
super(null,inputs,outputs);
}
public BaseDynamicCustomReduction(INDArray[] inputs, INDArray[] outputs,boolean keepDims) {
this(inputs,outputs,keepDims,null);
}
public BaseDynamicCustomReduction(INDArray[] inputs, INDArray[] outputs,boolean keepDims,int[] dimensions) {
this(inputs,outputs);
this.keepDims = keepDims;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(INDArray[] inputs,boolean keepDims,int[] dimensions) {
this(inputs,null,keepDims,dimensions);
}
public BaseDynamicCustomReduction(boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(SameDiff sameDiff, SDVariable arg, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(sameDiff, arg);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(SameDiff sameDiff, SDVariable[] args, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(sameDiff, args);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, SameDiff sameDiff, SDVariable[] args, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName, sameDiff, args);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, INDArray input, INDArray output, List tArguments, int[] iArguments, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName, input, output, tArguments, iArguments);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, INDArray[] inputs, INDArray[] outputs, List tArguments, int[] iArguments, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName, inputs, outputs, tArguments, iArguments);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, INDArray[] inputs, INDArray[] outputs, List tArguments, List iArguments, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName, inputs, outputs, tArguments, iArguments);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(INDArray[] inputs, INDArray[] outputs, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(inputs, outputs);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, INDArray[] inputs, INDArray[] outputs, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName, inputs, outputs);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, SameDiff sameDiff, SDVariable[] args, boolean inPlace, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName, sameDiff, args, inPlace);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(SameDiff sameDiff, SDVariable[] args, boolean inPlace, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(sameDiff, args, inPlace);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(String opName, boolean keepDims, boolean isComplex, boolean isEmptyReduce, int[] dimensions) {
super(opName);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.isEmptyReduce = isEmptyReduce;
this.dimensions = dimensions;
addArgs();
}
public BaseDynamicCustomReduction(INDArray[] input, INDArray[] output, boolean keepDims, boolean isComplex, int[] dimensions) {
super(null,input,output);
this.keepDims = keepDims;
this.isComplex = isComplex;
this.dimensions = dimensions;
addArgs();
}
protected void addArgs() {
addBArgument(keepDims);
if(dimensions != null) {
for(int i = 0; i < dimensions.length; i++) {
addIArgument(dimensions[i]);
}
}
}
/**
* Calculate the data types for the output arrays.
* Though datatypes can also be inferred from {@link #calculateOutputShape()}, this method differs in that it does not
* require the input arrays to be populated.
* This is important as it allows us to do greedy datatype inference for the entire net - even if arrays are not
* available.
*
* @param dataTypes The data types of the inputs
* @return The data types of the outputs
*/
@Override
public List calculateOutputDataTypes(List dataTypes){
if(!dArguments.isEmpty())
return Arrays.asList(dArguments.get(0));
return Arrays.asList(dataTypes.get(0));
}
public abstract String opName();
}
