blob: 9394dc4af89e945c33212bb724168dc4aeabd081 [file] [log] [blame]
/*
* Copyright (c) 2013 DataTorrent, Inc. ALL Rights Reserved.
*
* Licensed 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.datatorrent.lib.statistics;
import com.datatorrent.api.Context.OperatorContext;
import com.datatorrent.api.annotation.InputPortFieldAnnotation;
import com.datatorrent.api.annotation.OperatorAnnotation;
import com.datatorrent.api.annotation.OutputPortFieldAnnotation;
import com.datatorrent.api.DefaultInputPort;
import com.datatorrent.api.DefaultOutputPort;
import com.datatorrent.lib.util.BaseNumberValueOperator;
/**
* An implementation of BaseOperator that computes weighted mean of incoming data. <br>
* <p>
* <b>Input Port(s) : </b><br>
* <b>data : </b> Data values input port. <br>
* <b>weight : </b> Current input data weight. <br>
* <br>
* <b>Output Port(s) : </b> <br>
* <b>mean : </b>Weighted mean output port. <br>
* <br>
* <b>StateFull : Yes</b>, value are aggregated over application window. <br>
* <b>Partitions : No</b>, no will yeild wrong results. <br>
* <br>
* @displayName Weighted Mean
* @category Statistics
* @tags numeric, math, calculation, sum, count, mean operator, average
* @since 0.3.4
*/
@OperatorAnnotation(partitionable = false)
public class WeightedMeanOperator<V extends Number> extends BaseNumberValueOperator<V>
{
// aggregate weighted sum
private double weightedSum;
// aggregate weighted count
private double weightedCount;
// current input weight
private double currentWeight;
/**
* Input data port that takes a number.
*/
@InputPortFieldAnnotation(name = "data")
public final transient DefaultInputPort<V> data = new DefaultInputPort<V>()
{
/**
* Computes sum and count with each tuple
*/
@Override
public void process(V tuple)
{
weightedSum += currentWeight * tuple.doubleValue();
weightedCount += currentWeight;
}
};
/**
* Input weight port that takes a number.
*/
@InputPortFieldAnnotation(name = "weight")
public final transient DefaultInputPort<V> weight = new DefaultInputPort<V>()
{
/**
* Computes sum and count with each tuple
*/
@Override
public void process(V tuple)
{
if (tuple.doubleValue() != 0.0) currentWeight = tuple.doubleValue();
}
};
/**
* Output port that emits weighted mean.
*/
@OutputPortFieldAnnotation(name = "mean")
public final transient DefaultOutputPort<V> mean = new DefaultOutputPort<V>();
@Override
public void setup(OperatorContext arg0)
{
currentWeight = 1.0;
}
@Override
public void endWindow()
{
if (weightedCount != 0.0) {
mean.emit(getAverage());
}
weightedSum = 0.0;
weightedCount = 0.0;
}
/**
* Calculate average based on number type.
*/
@SuppressWarnings("unchecked")
public V getAverage()
{
if (weightedSum == 0) {
return null;
}
V num = getValue(weightedSum);
Number val;
switch (getType()) {
case DOUBLE:
val = new Double(num.doubleValue() / weightedCount);
break;
case INTEGER:
int icount = (int) (num.intValue() / weightedCount);
val = new Integer(icount);
break;
case FLOAT:
val = new Float(num.floatValue() / weightedCount);
break;
case LONG:
val = new Long((long) (num.longValue() / weightedCount));
break;
case SHORT:
short scount = (short) (num.shortValue() / weightedCount);
val = new Short(scount);
break;
default:
val = new Double(num.doubleValue() / weightedCount);
break;
}
return (V) val;
}
}