/// <summary>
/// Reads labeled feature values from a file with the specified name.
/// </summary>
/// <param name="fileName">The file name.</param>
/// <param name="labelSet">An optional set of labels.</param>
/// <param name="featureSet">An optional set of features.</param>
/// <returns>A list of labeled feature values.</returns>
public static IList <LabeledFeatureValues> LoadLabeledFeatureValues(
string fileName,
IndexedSet <string> labelSet = null,
IndexedSet <string> featureSet = null)
{
if (string.IsNullOrWhiteSpace(fileName))
{
throw new ArgumentException("The name of the file must not be null or whitespace.", "fileName");
}
var labeledFeatureValues = new List <LabeledFeatureValues>();
var labelDictionary = labelSet ?? new IndexedSet <string>();
var featureDictionary = featureSet ?? new IndexedSet <string>();
var parsingContext = new FileParsingContext(fileName);
using (var reader = new StreamReader(fileName))
{
string line;
while ((line = reader.ReadLine()) != null)
{
if (string.IsNullOrWhiteSpace(line))
{
continue;
}
line = line.Trim();
if (line.StartsWith("#") || line.StartsWith("//") || line.StartsWith("%"))
{
continue;
}
try
{
labeledFeatureValues.Add(LabeledFeatureValues.Parse(line, labelDictionary, featureDictionary));
}
catch (Exception e)
{
parsingContext.RaiseError("{0}", e.Message);
}
}
}
return(labeledFeatureValues);
}
/// <summary>
/// Parses a specified string for a label and its feature values.
/// </summary>
/// <param name="specification">The string to be parsed for a label and its feature values.</param>
/// <param name="labelSet">An optional bidirectional mapping from class labels to class indexes.</param>
/// <param name="featureSet">An optional bidirectional mapping from feature names to feature indexes.</param>
/// <returns>Labeled feature values equivalent to the given string specification.</returns>
public static LabeledFeatureValues Parse(
string specification,
IndexedSet <string> labelSet = null,
IndexedSet <string> featureSet = null)
{
var labeledFeatureValues = new LabeledFeatureValues();
if (featureSet != null)
{
labeledFeatureValues.FeatureSet = featureSet;
}
// Parse label distribution
labeledFeatureValues.LabelDistribution = LabelDistribution.Parse(specification, labelSet, out specification);
// Parse feature values
labeledFeatureValues.ParseFeatures(specification);
return(labeledFeatureValues);
}
