public TransformApplierParams(TextTransform parent)
{
var host = parent._host;
host.Check(Enum.IsDefined(typeof(Language), parent.AdvancedSettings.TextLanguage));
host.Check(Enum.IsDefined(typeof(CaseNormalizationMode), parent.AdvancedSettings.TextCase));
WordExtractorFactory = parent._wordFeatureExtractor?.CreateComponent(host, parent._dictionary);
CharExtractorFactory = parent._charFeatureExtractor?.CreateComponent(host, parent._dictionary);
VectorNormalizer = parent.AdvancedSettings.VectorNormalizer;
Language = parent.AdvancedSettings.TextLanguage;
StopWordsRemover = parent._stopWordsRemover;
TextCase = parent.AdvancedSettings.TextCase;
KeepDiacritics = parent.AdvancedSettings.KeepDiacritics;
KeepPunctuations = parent.AdvancedSettings.KeepPunctuations;
KeepNumbers = parent.AdvancedSettings.KeepNumbers;
OutputTextTokens = parent.AdvancedSettings.OutputTokens;
Dictionary = parent._dictionary;
}
public static IDataTransform Create(IHostEnvironment env, Arguments args, IDataView data)
{
Action <Settings> settings = s =>
{
s.TextLanguage = args.Language;
s.TextCase = args.TextCase;
s.KeepDiacritics = args.KeepDiacritics;
s.KeepPunctuations = args.KeepPunctuations;
s.KeepNumbers = args.KeepNumbers;
s.OutputTokens = args.OutputTokens;
s.VectorNormalizer = args.VectorNormalizer;
};
var estimator = new TextTransform(env, args.Column.Source ?? new[] { args.Column.Name }, args.Column.Name, settings);
estimator._stopWordsRemover = args.StopWordsRemover;
estimator._dictionary = args.Dictionary;
estimator._wordFeatureExtractor = args.WordFeatureExtractor;
estimator._charFeatureExtractor = args.CharFeatureExtractor;
return(estimator.Fit(data).Transform(data) as IDataTransform);
}