com.simiacryptus.mindseye.layers.java.NthPowerActivationLayer Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of mindseye-java Show documentation
Show all versions of mindseye-java Show documentation
Pure Java Neural Networks Components
The newest version!
/*
* Copyright (c) 2019 by Andrew Charneski.
*
* The author licenses this file to you under the
* Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance
* with the License. You may obtain a copy
* of the License at
*
* http://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.
*/
package com.simiacryptus.mindseye.layers.java;
import com.google.gson.JsonObject;
import com.simiacryptus.mindseye.lang.*;
import com.simiacryptus.ref.lang.RefIgnore;
import com.simiacryptus.ref.lang.RefUtil;
import com.simiacryptus.ref.wrappers.RefArrays;
import com.simiacryptus.ref.wrappers.RefIntStream;
import com.simiacryptus.ref.wrappers.RefList;
import org.jetbrains.annotations.NotNull;
import javax.annotation.Nonnull;
import javax.annotation.Nullable;
import java.util.Map;
import java.util.UUID;
import java.util.function.IntFunction;
/**
* The type Nth power activation layer.
*/
@SuppressWarnings("serial")
public final class NthPowerActivationLayer extends LayerBase {
private double power = 1.0;
/**
* Instantiates a new Nth power activation layer.
*/
public NthPowerActivationLayer() {
}
/**
* Instantiates a new Nth power activation layer.
*
* @param id the id
*/
protected NthPowerActivationLayer(@Nonnull final JsonObject id) {
super(id);
power = id.get("power").getAsDouble();
}
/**
* Instantiates a new Nth power activation layer.
*
* @param power the power
*/
public NthPowerActivationLayer(double power) {
this.power = power;
}
/**
* Gets power.
*
* @return the power
*/
public double getPower() {
return power;
}
/**
* Sets power.
*
* @param power the power
*/
public void setPower(double power) {
this.power = power;
}
/**
* From json nth power activation layer.
*
* @param json the json
* @param rs the rs
* @return the nth power activation layer
*/
@Nonnull
@SuppressWarnings("unused")
public static NthPowerActivationLayer fromJson(@Nonnull final JsonObject json, Map rs) {
return new NthPowerActivationLayer(json);
}
private static void nthPower(final double power, @Nonnull final Tensor input, final double[] inputData,
final double[] gradientData, final double[] outputData) {
for (int i = 0; i < input.length(); i++) {
final double x = inputData[i];
final boolean isZero = Math.abs(x) < 1e-20;
double d = isZero ? 0.0 : power * Math.pow(x, power - 1);
double f = isZero ? 0.0 : Math.pow(x, power);
if (!Double.isFinite(d)) {
d = 0.0;
}
if (!Double.isFinite(f)) {
f = 0.0;
}
gradientData[i] = d;
outputData[i] = f;
}
input.freeRef();
}
private static void square(@Nonnull final Tensor input, final double[] inputData, final double[] gradientData,
final double[] outputData) {
for (int i = 0; i < input.length(); i++) {
final double x = inputData[i];
gradientData[i] = 2 * x;
outputData[i] = x * x;
}
input.freeRef();
}
private static void squareRoot(@Nonnull final Tensor input, final double[] inputData, final double[] gradientData,
final double[] outputData) {
for (int i = 0; i < input.length(); i++) {
final double x = inputData[i];
final boolean isZero = Math.abs(x) < 1e-20;
final double power = 0.5;
final double v = Math.pow(x, power);
double d = isZero ? 0.0 : power / v;
double f = isZero ? 0.0 : v;
if (!Double.isFinite(d)) {
d = 0.0;
}
if (!Double.isFinite(f)) {
f = 0.0;
}
gradientData[i] = d;
outputData[i] = f;
}
input.freeRef();
}
private static void unity(@Nonnull final Tensor input, final double[] gradientData,
final double[] outputData) {
for (int i = 0; i < input.length(); i++) {
gradientData[i] = 0;
outputData[i] = 1;
}
input.freeRef();
}
@Nonnull
@Override
public Result eval(@Nonnull final Result... inObj) {
final Result in0 = inObj[0].addRef();
RefUtil.freeRef(inObj);
TensorList inData = in0.getData();
final int itemCnt = inData.length();
assert 0 < itemCnt;
@Nonnull final Tensor inputGradientA[] = new Tensor[itemCnt];
TensorArray data = fwd(inData, itemCnt, inputGradientA);
boolean alive = 0.0 != power && in0.isAlive();
Result.Accumulator accumulator = new Accumulator(inputGradientA, itemCnt, in0.getAccumulator(), in0.isAlive());
in0.freeRef();
return new Result(data, accumulator, alive);
}
@Nonnull
@Override
public JsonObject getJson(Map resources, DataSerializer dataSerializer) {
@Nonnull final JsonObject json = super.getJsonStub();
json.addProperty("power", power);
return json;
}
@Nonnull
@Override
public RefList state() {
return RefArrays.asList();
}
public @SuppressWarnings("unused")
void _free() {
super._free();
}
@Nonnull
public @Override
@SuppressWarnings("unused")
NthPowerActivationLayer addRef() {
return (NthPowerActivationLayer) super.addRef();
}
@NotNull
private TensorArray fwd(TensorList inData, int itemCnt, @RefIgnore Tensor[] inputGradientA) {
return new TensorArray(RefIntStream.range(0, itemCnt).parallel()
.mapToObj(RefUtil.wrapInterface((IntFunction) dataIndex -> {
@Nullable final Tensor input = inData.get(dataIndex);
@Nonnull final Tensor output = new Tensor(inData.getDimensions());
@Nonnull final Tensor gradient = new Tensor(input.length());
@Nullable final double[] inputData = input.getData();
@Nullable final double[] gradientData = gradient.getData();
@Nullable final double[] outputData = output.getData();
RefUtil.set(inputGradientA, dataIndex, gradient);
if (power == 2) {
NthPowerActivationLayer.square(input.addRef(), inputData, gradientData,
outputData);
} else if (power == 0.5) {
NthPowerActivationLayer.squareRoot(input.addRef(), inputData, gradientData,
outputData);
} else if (power == 0.0) {
NthPowerActivationLayer.unity(input.addRef(), gradientData,
outputData);
} else {
NthPowerActivationLayer.nthPower(power, input.addRef(), inputData, gradientData,
outputData);
}
input.freeRef();
return output;
}, inData)).toArray(Tensor[]::new));
}
private static class Accumulator extends Result.Accumulator {
private final Tensor[] inputGradientA;
private final int itemCnt;
private Result.Accumulator accumulator;
private boolean alive;
/**
* Instantiates a new Accumulator.
*
* @param inputGradientA the input gradient a
* @param itemCnt the item cnt
* @param accumulator the accumulator
* @param alive the alive
*/
public Accumulator(Tensor[] inputGradientA, int itemCnt, Result.Accumulator accumulator, boolean alive) {
this.inputGradientA = inputGradientA;
this.itemCnt = itemCnt;
this.accumulator = accumulator;
this.alive = alive;
}
@Override
public void accept(@Nullable DeltaSet buffer, @Nonnull TensorList data) {
if (alive) {
this.accumulator.accept(buffer, new TensorArray(RefIntStream.range(0, itemCnt).parallel()
.mapToObj(RefUtil.wrapInterface((IntFunction) dataIndex -> {
@Nonnull final Tensor passback = new Tensor(data.getDimensions());
@Nullable final Tensor tensor = data.get(dataIndex);
@Nullable
double[] tensorData = tensor.getData();
@Nullable final double[] gradientData = inputGradientA[dataIndex].getData();
RefIntStream.range(0, passback.length()).forEach(RefUtil.wrapInterface(i -> {
final double v = gradientData[i];
if (Double.isFinite(v)) {
passback.set(i, tensorData[i] * v);
}
}, passback.addRef()));
tensor.freeRef();
return passback;
}, data)).toArray(Tensor[]::new)));
} else {
data.freeRef();
if (null != buffer)
buffer.freeRef();
}
}
public @SuppressWarnings("unused")
void _free() {
super._free();
accumulator.freeRef();
RefUtil.freeRef(inputGradientA);
}
}
}