edu.berkeley.nlp.util.VariableSizeHistogram Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of berkeleyparser Show documentation

The Berkeley parser analyzes the grammatical structure of natural language using probabilistic context-free grammars (PCFGs).

The newest version!

package edu.berkeley.nlp.util;

import java.text.NumberFormat;
import java.util.SortedSet;
import java.util.Map.Entry;

public class VariableSizeHistogram
{

	private SortedSet startPoints;

	private Counter data;

	private boolean intKeys;

	public VariableSizeHistogram(SortedSet startPoints)
	{
		this(startPoints, true);
	}

	public VariableSizeHistogram(SortedSet startPoints, boolean intKeys)
	{
	
		this.startPoints = startPoints;
		this.intKeys = intKeys;
		

		data = new Counter();
	}

	public void add(double key, double value)
	{
		SortedSet tailSet = startPoints.tailSet(key);
		Double startPoint = tailSet.isEmpty() ? startPoints.last() : tailSet.first();
		data.incrementCount(startPoint, value);
	}

	public void addAll(Counter counter)
	{
		for (Entry entry : counter.entrySet())
		{
			add(entry.getKey().doubleValue(), entry.getValue());
		}
	}

	@Override
	public String toString()
	{

		StringBuilder sb = new StringBuilder("[");

		NumberFormat f = NumberFormat.getInstance();
		f.setMaximumFractionDigits(5);
		Number previous = null;
		for (Number element : startPoints)
		{
			if (previous != null)
			{
				Number rangeEnd = null;
				if (intKeys)
				{
					int num = element.intValue() - 1;
					int prev = previous.intValue();
					if (num != prev)
						rangeEnd = num;
				}
				else
					rangeEnd = element;

				sb.append(f.format(previous) + (rangeEnd == null ? "" : "-" + f.format(rangeEnd)));
				sb.append(" : ");
				sb.append(f.format(data.getCount(element)));
				sb.append(", ");

			}
			previous = element;
		}

		sb.append(f.format(previous) + "-");
		sb.append(" : ");
		sb.append(f.format(data.getCount(previous)));

		sb.append("]");
		return sb.toString();
	}
}