org.nd4j.linalg.api.ops.random.impl.BernoulliDistribution Maven / Gradle / Ivy
/*******************************************************************************
* Copyright (c) 2015-2018 Skymind, Inc.
*
* 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.
*
* 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.random.impl;
import lombok.NonNull;
import org.nd4j.autodiff.samediff.SDVariable;
import org.nd4j.autodiff.samediff.SameDiff;
import org.nd4j.base.Preconditions;
import org.nd4j.imports.NoOpNameFoundException;
import org.nd4j.linalg.api.buffer.DataType;
import org.nd4j.linalg.api.ndarray.INDArray;
import org.nd4j.linalg.api.ops.random.BaseRandomOp;
import org.nd4j.linalg.exception.ND4JIllegalStateException;
import java.util.Collections;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;
/**
* BernoulliDistribution implementation
*
* @author [email protected]
*/
public class BernoulliDistribution extends BaseRandomOp {
private double prob;
public BernoulliDistribution(SameDiff sd, double prob, long[] shape){
super(sd, shape);
this.prob = prob;
this.extraArgs = new Object[] {this.prob};
}
public BernoulliDistribution() {
super();
}
/**
* This op fills Z with bernoulli trial results, so 0, or 1, depending by common probability
* @param z
*/
public BernoulliDistribution(@NonNull INDArray z, double prob) {
super(null, null, z);
this.prob = prob;
this.extraArgs = new Object[] {this.prob};
}
/**
* This op fills Z with bernoulli trial results, so 0, or 1, each element will have it's own success probability defined in prob array
* @param prob array with probabilities
* @param z
*/
public BernoulliDistribution(@NonNull INDArray z, @NonNull INDArray prob) {
super(prob, null, z);
if (prob.elementWiseStride() != 1)
throw new ND4JIllegalStateException("Probabilities should have ElementWiseStride of 1");
if (prob.length() != z.length())
throw new ND4JIllegalStateException("Length of probabilities array [" + prob.length()
+ "] doesn't match length of output array [" + z.length() + "]");
this.prob = 0.0;
this.extraArgs = new Object[] {this.prob};
}
@Override
public int opNum() {
return 7;
}
@Override
public String opName() {
return "distribution_bernoulli";
}
@Override
public String onnxName() {
throw new NoOpNameFoundException("No onnx op opName found for " + opName());
}
@Override
public String tensorflowName() {
throw new NoOpNameFoundException("No tensorflow op opName found for " + opName());
}
@Override
public List doDiff(List f1) {
return Collections.emptyList(); //No SDVariable args
}
@Override
public List calculateOutputDataTypes(List inputDataTypes){
Preconditions.checkState(inputDataTypes == null || inputDataTypes.isEmpty(), "Expected no input datatypes (no args) for %s, got %s", getClass(), inputDataTypes);
//Input data type specifies the shape; output data type should be any float
//TODO MAKE CONFIGUREABLE - https://github.com/deeplearning4j/deeplearning4j/issues/6854
return Collections.singletonList(DataType.DOUBLE);
}
}