org.nd4j.linalg.dataset.api.preprocessor.serializer.MultiStandardizeSerializerStrategy 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.dataset.api.preprocessor.serializer;
import lombok.NonNull;
import org.nd4j.linalg.dataset.api.preprocessor.MultiNormalizerStandardize;
import org.nd4j.linalg.dataset.api.preprocessor.stats.DistributionStats;
import org.nd4j.linalg.factory.Nd4j;
import java.io.*;
import java.util.ArrayList;
import java.util.List;
public class MultiStandardizeSerializerStrategy implements NormalizerSerializerStrategy {
/**
* Serialize a MultiNormalizerStandardize to a output stream
*
* @param normalizer the normalizer
* @param stream the output stream to write to
* @throws IOException
*/
public void write(@NonNull MultiNormalizerStandardize normalizer, @NonNull OutputStream stream) throws IOException {
try (DataOutputStream dos = new DataOutputStream(stream)) {
dos.writeBoolean(normalizer.isFitLabel());
dos.writeInt(normalizer.numInputs());
dos.writeInt(normalizer.isFitLabel() ? normalizer.numOutputs() : -1);
for (int i = 0; i < normalizer.numInputs(); i++) {
Nd4j.write(normalizer.getFeatureMean(i), dos);
Nd4j.write(normalizer.getFeatureStd(i), dos);
}
if (normalizer.isFitLabel()) {
for (int i = 0; i < normalizer.numOutputs(); i++) {
Nd4j.write(normalizer.getLabelMean(i), dos);
Nd4j.write(normalizer.getLabelStd(i), dos);
}
}
dos.flush();
}
}
/**
* Restore a MultiNormalizerStandardize that was previously serialized by this strategy
*
* @param stream the input stream to restore from
* @return the restored MultiNormalizerStandardize
* @throws IOException
*/
public MultiNormalizerStandardize restore(@NonNull InputStream stream) throws IOException {
DataInputStream dis = new DataInputStream(stream);
boolean fitLabels = dis.readBoolean();
int numInputs = dis.readInt();
int numOutputs = dis.readInt();
MultiNormalizerStandardize result = new MultiNormalizerStandardize();
result.fitLabel(fitLabels);
List featureStats = new ArrayList<>();
for (int i = 0; i < numInputs; i++) {
featureStats.add(new DistributionStats(Nd4j.read(dis), Nd4j.read(dis)));
}
result.setFeatureStats(featureStats);
if (fitLabels) {
List labelStats = new ArrayList<>();
for (int i = 0; i < numOutputs; i++) {
labelStats.add(new DistributionStats(Nd4j.read(dis), Nd4j.read(dis)));
}
result.setLabelStats(labelStats);
}
return result;
}
@Override
public NormalizerType getSupportedType() {
return NormalizerType.MULTI_STANDARDIZE;
}
}