Examples with ColumnRearranger org.knime.core.data.container.ColumnRearranger used on opensource projects

Example 46 with ColumnRearranger

use of org.knime.core.data.container.ColumnRearranger in project knime-core by knime.

the class MLPPredictorNodeModel method execute.

/**
 * {@inheritDoc}
 */
@Override
public PortObject[] execute(final PortObject[] inData, final ExecutionContext exec) throws Exception {
    BufferedDataTable testdata = (BufferedDataTable) inData[1];
    PMMLPortObject pmmlPort = (PMMLPortObject) inData[0];
    List<Node> models = pmmlPort.getPMMLValue().getModels(PMMLModelType.NeuralNetwork);
    if (models.isEmpty()) {
        String msg = "Neural network evaluation failed: " + "No neural network model found.";
        LOGGER.error(msg);
        throw new RuntimeException(msg);
    }
    PMMLNeuralNetworkTranslator trans = new PMMLNeuralNetworkTranslator();
    pmmlPort.initializeModelTranslator(trans);
    m_mlp = trans.getMLP();
    m_columns = getLearningColumnIndices(testdata.getDataTableSpec(), pmmlPort.getSpec());
    DataColumnSpec targetCol = pmmlPort.getSpec().getTargetCols().iterator().next();
    MLPClassificationFactory mymlp;
    /*
         * Regression
         */
    if (m_mlp.getMode() == MultiLayerPerceptron.REGRESSION_MODE) {
        mymlp = new MLPClassificationFactory(true, m_columns, targetCol);
    } else if (m_mlp.getMode() == MultiLayerPerceptron.CLASSIFICATION_MODE) {
        /*
             * Classification
             */
        mymlp = new MLPClassificationFactory(false, m_columns, targetCol);
    } else {
        throw new Exception("Unsupported Mode: " + m_mlp.getMode());
    }
    ColumnRearranger colre = new ColumnRearranger(testdata.getDataTableSpec());
    colre.append(mymlp);
    BufferedDataTable bdt = exec.createColumnRearrangeTable(testdata, colre, exec);
    return new BufferedDataTable[] { bdt };
}

Example 47 with ColumnRearranger

use of org.knime.core.data.container.ColumnRearranger in project knime-core by knime.

the class RegressionPredictorNodeModel method createRearranger.

private ColumnRearranger createRearranger(final PMMLGeneralRegressionContent content, final PMMLPortObjectSpec pmmlSpec, final DataTableSpec inDataSpec) throws InvalidSettingsException {
    if (content == null) {
        throw new InvalidSettingsException("No input");
    }
    // the predictor can only predict linear regression models
    if (!(content.getModelType().equals(ModelType.multinomialLogistic) || content.getModelType().equals(ModelType.generalLinear))) {
        throw new InvalidSettingsException("Model Type: " + content.getModelType() + " is not supported.");
    }
    if (content.getModelType().equals(ModelType.generalLinear) && !content.getFunctionName().equals(FunctionName.regression)) {
        throw new InvalidSettingsException("Function Name: " + content.getFunctionName() + " is not supported for linear regression.");
    }
    if (content.getModelType().equals(ModelType.multinomialLogistic) && !content.getFunctionName().equals(FunctionName.classification)) {
        throw new InvalidSettingsException("Function Name: " + content.getFunctionName() + " is not supported for logistic regression.");
    }
    // are nominal values
    for (PMMLPredictor factor : content.getFactorList()) {
        DataColumnSpec columnSpec = inDataSpec.getColumnSpec(factor.getName());
        if (null == columnSpec) {
            throw new InvalidSettingsException("The column \"" + factor.getName() + "\" is in the model but not in given table.");
        }
        if (!columnSpec.getType().isCompatible(NominalValue.class)) {
            throw new InvalidSettingsException("The column \"" + factor.getName() + "\" is supposed to be nominal.");
        }
    }
    // check if all covariates are in the given data table and that they
    // are numeric values
    Pattern pattern = Pattern.compile("(.*)\\[\\d+\\]");
    for (PMMLPredictor covariate : content.getCovariateList()) {
        DataColumnSpec columnSpec = inDataSpec.getColumnSpec(covariate.getName());
        if (null == columnSpec) {
            Matcher matcher = pattern.matcher(covariate.getName());
            boolean found = matcher.matches();
            columnSpec = inDataSpec.getColumnSpec(matcher.group(1));
            found = found && null != columnSpec;
            if (!found) {
                throw new InvalidSettingsException("The column \"" + covariate.getName() + "\" is in the model but not in given table.");
            }
        }
        if (columnSpec != null && !columnSpec.getType().isCompatible(DoubleValue.class) && !(content.getVectorLengths().containsKey(columnSpec.getName()) && ((columnSpec.getType().isCollectionType() && columnSpec.getType().getCollectionElementType().isCompatible(DoubleValue.class)) || columnSpec.getType().isCompatible(BitVectorValue.class) || columnSpec.getType().isCompatible(ByteVectorValue.class)))) {
            throw new InvalidSettingsException("The column \"" + covariate.getName() + "\" is supposed to be numeric.");
        }
    }
    ColumnRearranger c = new ColumnRearranger(inDataSpec);
    if (content.getModelType().equals(ModelType.generalLinear)) {
        c.append(new LinReg2Predictor(content, inDataSpec, pmmlSpec, pmmlSpec.getTargetFields().get(0), m_settings));
    } else {
        c.append(new LogRegPredictor(content, inDataSpec, pmmlSpec, pmmlSpec.getTargetFields().get(0), m_settings));
    }
    return c;
}

Example 48 with ColumnRearranger

use of org.knime.core.data.container.ColumnRearranger in project knime-core by knime.

the class RegressionPredictorNodeModel method execute.

/**
 * {@inheritDoc}
 */
@Override
public PortObject[] execute(final PortObject[] inData, final ExecutionContext exec) throws Exception {
    PMMLPortObject port = (PMMLPortObject) inData[0];
    List<Node> models = port.getPMMLValue().getModels(PMMLModelType.GeneralRegressionModel);
    if (models.isEmpty()) {
        LOGGER.warn("No regression models in the input PMML.");
        @SuppressWarnings("deprecation") org.knime.base.node.mine.regression.predict.RegressionPredictorNodeModel regrPredictor = new org.knime.base.node.mine.regression.predict.RegressionPredictorNodeModel();
        @SuppressWarnings("deprecation") PortObject[] regrPredOut = regrPredictor.execute(inData, exec);
        if (regrPredOut.length > 0 && regrPredOut[0] instanceof BufferedDataTable) {
            BufferedDataTable regrPredOutTable = (BufferedDataTable) regrPredOut[0];
            // replace name of prediction column (the last column of regrPredOutTable)
            return new PortObject[] { adjustSpecOfRegressionPredictorTable(regrPredOutTable, inData, exec) };
        } else {
            return regrPredOut;
        }
    }
    PMMLGeneralRegressionTranslator trans = new PMMLGeneralRegressionTranslator();
    port.initializeModelTranslator(trans);
    BufferedDataTable data = (BufferedDataTable) inData[1];
    DataTableSpec spec = data.getDataTableSpec();
    ColumnRearranger c = createRearranger(trans.getContent(), port.getSpec(), spec);
    BufferedDataTable out = exec.createColumnRearrangeTable(data, c, exec);
    return new BufferedDataTable[] { out };
}

Example 49 with ColumnRearranger

use of org.knime.core.data.container.ColumnRearranger in project knime-core by knime.

the class RegressionPredictorNodeModel method configure.

/**
 * {@inheritDoc}
 */
@Override
protected PortObjectSpec[] configure(final PortObjectSpec[] inSpecs) throws InvalidSettingsException {
    PMMLPortObjectSpec regModelSpec = (PMMLPortObjectSpec) inSpecs[0];
    DataTableSpec dataSpec = (DataTableSpec) inSpecs[1];
    if (dataSpec == null || regModelSpec == null) {
        throw new InvalidSettingsException("No input specification available");
    }
    if (regModelSpec.getTargetCols().get(0).getType().isCompatible(DoubleValue.class) && m_settings.getIncludeProbabilities()) {
        setWarningMessage("The option \"Append columns with predicted probabilities\"" + " has only an effect for nominal targets");
    }
    if (null != RegressionPredictorCellFactory.createColumnSpec(regModelSpec, dataSpec, m_settings)) {
        ColumnRearranger c = new ColumnRearranger(dataSpec);
        c.append(new RegressionPredictorCellFactory(regModelSpec, dataSpec, m_settings) {

            @Override
            public DataCell[] getCells(final DataRow row) {
                // not called during configure
                return null;
            }
        });
        DataTableSpec outSpec = c.createSpec();
        return new DataTableSpec[] { outSpec };
    } else {
        return null;
    }
}

Example 50 with ColumnRearranger

use of org.knime.core.data.container.ColumnRearranger in project knime-core by knime.

the class BitVectorGeneratorNodeModel method configure.

/**
 * Assume to get numeric data only. Output is one column of type BitVector.
 *
 * {@inheritDoc}
 */
@Override
protected DataTableSpec[] configure(final DataTableSpec[] inSpecs) throws InvalidSettingsException {
    DataTableSpec spec = inSpecs[0];
    if (!m_fromString) {
        // check if there is at least one numeric column selected
        if (m_includedColumns.isEnabled() && m_includedColumns.getIncludeList().isEmpty()) {
            // the includeColumns model cannot be empty
            // through the dialog (see #validateSettings)
            // only case where !m_fromString and includeColumns evaluates
            // to true is for old workflows.
            // For backward compatiblity include all numeric columns
            // which was the behavior before 2.0
            List<String> allNumericColumns = new ArrayList<String>();
            for (DataColumnSpec colSpec : spec) {
                if (colSpec.getType().isCompatible(DoubleValue.class)) {
                    allNumericColumns.add(colSpec.getName());
                }
            }
            m_includedColumns.setIncludeList(allNumericColumns);
            m_loadedSettingsDontHaveIncludeColumns = false;
        }
        for (String inclColName : m_includedColumns.getIncludeList()) {
            DataColumnSpec colSpec = spec.getColumnSpec(inclColName);
            if (colSpec == null) {
                throw new InvalidSettingsException("Column " + inclColName + " not found in input table. " + "Please re-configure the node.");
            }
            if (!colSpec.getType().isCompatible(DoubleValue.class)) {
                throw new InvalidSettingsException("Column " + inclColName + " is not a numeric column");
            }
        }
    } else {
        // parse from string column
        if (m_stringColumn.getStringValue() == null) {
            throw new InvalidSettingsException("No string column selected. " + "Please (re-)configure the node.");
        }
        // -> check if selected column is a string column
        if (!spec.containsName(m_stringColumn.getStringValue()) || !(spec.getColumnSpec(m_stringColumn.getStringValue()).getType().isCompatible(StringValue.class))) {
            throw new InvalidSettingsException("Selected string column " + m_stringColumn.getStringValue() + " not in the input table");
        }
    }
    if (m_fromString) {
        int stringColIdx = inSpecs[0].findColumnIndex(m_stringColumn.getStringValue());
        ColumnRearranger c = createColumnRearranger(inSpecs[0], stringColIdx);
        return new DataTableSpec[] { c.createSpec() };
    } else {
        // numeric input
        DataTableSpec newSpec;
        DataColumnSpec newColSpec = createNumericOutputSpec(spec);
        if (m_replace) {
            ColumnRearranger colR = new ColumnRearranger(spec);
            colR.remove(m_includedColumns.getIncludeList().toArray(new String[m_includedColumns.getIncludeList().size()]));
            newSpec = new DataTableSpec(colR.createSpec(), new DataTableSpec(newColSpec));
        } else {
            newSpec = new DataTableSpec(spec, new DataTableSpec(newColSpec));
        }
        return new DataTableSpec[] { newSpec };
    }
}