weka.core.converters.ConverterUtils Java Examples

/**
 * Returns the next dataset.
 *
 * @return      the next dataset, null in case of an error
 */
@Override
public Instances next() {
	Instances result;
	try {
		log("Loading: " + m_Datasets.get(m_Current));
		result = ConverterUtils.DataSource.read(m_Datasets.get(m_Current).getAbsolutePath());
		MLUtils.prepareData(result);
	}
	catch (Exception e) {
		result = null;
		handleException("Failed to load dataset: " + m_Datasets.get(m_Current), e);
	}
	m_Current++;
	return result;
}
 

/**
 * Saves the data to the specified file.
 *
 * @param file the file to save the data to
 * @param saver the saver to use, determines it automatically if null
 */
public void save(File file, AbstractFileSaver saver) {
	if (saver == null)
		saver = ConverterUtils.getSaverForFile(file);
	try {
		log(null, "Saving: " + file);
		saver.setInstances(m_Data);
		if ((saver.retrieveFile() == null) || !saver.retrieveFile().equals(file))
			saver.setFile(file);
		saver.writeBatch();
		m_CurrentFile = file;
		log(null, "Saved successfully: " + file);
	}
	catch (Exception e) {
		handleException(null, "Failed to save data to '" + file + "':", e);
		JOptionPane.showMessageDialog(
				this,
				"Failed to save dataset to '" + file + "':\n" + e,
				"Error saving",
				JOptionPane.ERROR_MESSAGE);
	}

	updateMenu();
}
 

public void testReuters() throws Exception {
  final String arffPath = "datasets/text/ReutersCorn-train.arff";
  ConverterUtils.DataSource ds = new ConverterUtils.DataSource(arffPath);
  final Instances data = ds.getDataSet();
  Dl4jStringToWord2Vec dl4jw2v = new Dl4jStringToWord2Vec();
  dl4jw2v.setInputFormat(data);
  Instances d = Filter.useFilter(data, dl4jw2v);
}
 

/**
 * loads the specified file
 *
 * @param filename the file to load
 * @param loaders optional varargs loader to use
 */
public void loadFile(String filename, AbstractFileLoader... loaders) {
	DataPanel panel;
	AbstractFileLoader loader;

	panel = new DataPanel(filename, loaders);
	panel.addChangeListener(this);
	m_TabbedPane.addTab(panel.getTitle(), panel);
	m_TabbedPane.setSelectedIndex(m_TabbedPane.getTabCount() - 1);
	if (loaders == null)
		loader = ConverterUtils.getLoaderForFile(filename);
	else
		loader = loaders[0];
	m_RecentFilesHandler.addRecentItem(new RecentFilesHandlerWithCommandline.Setup(new File(filename), loader));
}
 

/**
 * saves the current data into a file
 */
public void saveFile() {
	DataPanel panel;
	String filename;
	AbstractSaver saver;

	// no panel? -> exit
	panel = getCurrentPanel();
	if (panel == null) {
		return;
	}

	filename = panel.getFilename();

	if (filename.equals(DataPanel.TAB_INSTANCES)) {
		saveFileAs();
	} else {
		saver = ConverterUtils.getSaverForFile(filename);
		try {
			saver.setFile(new File(filename));
			saver.setInstances(panel.getInstances());
			saver.writeBatch();
			panel.setChanged(false);
			setCurrentFilename(filename);
		} catch (Exception e) {
			e.printStackTrace();
		}
	}
}
 

/**
 * loads the specified ARFF file
 *
 * @param filename the file to load
 * @param loaders optional varargs for a loader to use
 */
protected void loadFile(String filename, AbstractFileLoader... loaders) {
	AbstractFileLoader loader;
	Instances data;

	if (loaders == null || loaders.length == 0) {
		loader = ConverterUtils.getLoaderForFile(filename);
	} else {
		loader = loaders[0];
	}

	if (loader != null) {
		try {
			loader.setFile(new File(filename));
			data = loader.getDataSet();
			// fix class attributes definition in relation name if necessary
			MLUtils.fixRelationName(data);
			MLUtils.prepareData(data);
			setInstances(data);
		} catch (Exception e) {
			ComponentHelper
				.showMessageBox(null, "Error loading file...", e.toString(),
					JOptionPane.OK_CANCEL_OPTION, JOptionPane.ERROR_MESSAGE);
			System.out.println(e);
			setInstances(null);
		}
	}
}
 

/**
 * Called by JUnit before each test method. This implementation creates
 * the default filter to test and loads a test set of Instances.
 *
 * @throws Exception if an error occurs reading the example instances.
 */
protected void setUp() throws Exception {
  m_Filter             = getFilter();
  m_Instances          = ConverterUtils.DataSource.read("FilterTest.arff");
  m_OptionTester       = getOptionTester();
  m_GOETester          = getGOETester();
  m_FilteredClassifier = getFilteredClassifier();
  m_FilteredClassifier.setDoNotCheckForModifiedClassAttribute(true);
}
 

public void testReuters() throws Exception {
  final String arffPath = "datasets/text/ReutersCorn-train.arff";
  ConverterUtils.DataSource ds = new ConverterUtils.DataSource(arffPath);
  final Instances data = ds.getDataSet();
  Dl4jStringToWord2Vec dl4jw2v = new Dl4jStringToWord2Vec();
  dl4jw2v.setInputFormat(data);
  Instances d = Filter.useFilter(data, dl4jw2v);
}
 

/**
 * Creates neuroph dataset from arff weka file.
 *
 * @param filePath Path to the file.
 * @param numInputs Number of inputs.
 * @param numOutputs Number of outputs.
 * @return Neuroph dataset.
 */
public static DataSet createDataSetFromFile(String filePath, int numInputs, int numOutputs) {

    try {
        ConverterUtils.DataSource dataSource = new ConverterUtils.DataSource(filePath);

        Instances wekaDataset = dataSource.getDataSet();

        wekaDataset.setClassIndex(numInputs);

        DataSet neurophDataset = WekaDataSetConverter.convertWekaToNeurophDataset(wekaDataset, numInputs, numOutputs);

        return neurophDataset;

    } catch (Exception e) {
        return new DataSet(0);
    }
}
 

/**
 * Processes the commandline arguments.
 *
 * @param args the arguments
 */
public void processCommandLineArgs(String[] args) {
	if (args.length > 0)
		open(new File(args[0]), ConverterUtils.getLoaderForFile(args[0]));
}
 

/**
 * Loads the dataset from disk.
 *
 * @param file the dataset to load (e.g., "weka/classifiers/data/something.arff")
 * @throws Exception if loading fails, e.g., file does not exit
 */
public static Instances loadData(String file) throws Exception {
	return ConverterUtils.DataSource.read(file);
}