netkit.classifiers.io
Class ReadEstimateRainbow

java.lang.Object
  netkit.classifiers.io.ReadEstimateRainbow

All Implemented Interfaces:

ReadEstimate

public final class ReadEstimateRainbow
extends java.lang.Object
implements ReadEstimate

Reads in estimates of entities in a graph from a file where the format of each line is:

nodeID trueclass class:score ... class:score

Lines starting with '#' are ignored. This line format is equivalent to what you would see in the output from the Rainbow text classification package developed by Andrew McCallum.

Author:

Sofus A. Macskassy (sofmac@gmail.com)

Constructor Summary

ReadEstimateRainbow()

Method Summary

void	readEstimate(Graph graph, Estimate estimates, java.io.File input) Reads in a set of estimates from the given file, assuming that each line is of the form: nodeID trueclass class:score ...
Estimate	readEstimate(Graph graph, java.lang.String nodeType, AttributeCategorical attribute, java.io.File input) Create a new Estimate object and call the more general readEstimate method.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ReadEstimateRainbow

public ReadEstimateRainbow()

Method Detail

readEstimate

public Estimate readEstimate(Graph graph,
                             java.lang.String nodeType,
                             AttributeCategorical attribute,
                             java.io.File input)

Create a new Estimate object and call the more general readEstimate method.

Specified by:

readEstimate in interface ReadEstimate

Parameters:

graph -

nodeType -

attribute -

input -

Returns:

A new Estimate object filled in with Estimates from the input file

readEstimate

public void readEstimate(Graph graph,
                         Estimate estimates,
                         java.io.File input)

Reads in a set of estimates from the given file, assuming that each line is of the form:

nodeID trueclass class:score ... class:score

where all classes in the underlying categorical attribute will be given a score.

ASSUMPTION: The underlying categorical attribute has already been initialized with the valid names for the attributes. Otherwise exceptions will be thrown!

Specified by:

readEstimate in interface ReadEstimate

Parameters:

graph -

estimates -

input -