Examples with PMMLPortObjectSpecCreator org.knime.core.node.port.pmml.PMMLPortObjectSpecCreator used on opensource projects

Example 16 with PMMLPortObjectSpecCreator

use of org.knime.core.node.port.pmml.PMMLPortObjectSpecCreator in project knime-core by knime.

the class LinReg2Learner method init.

/**
 * Initialize instance and check if settings are consistent.
 */
private void init(final DataTableSpec inSpec, final PMMLPortObjectSpec pmmlSpec, final Set<String> exclude) throws InvalidSettingsException {
    m_warningMessage = null;
    // Auto configuration when target is not set
    if (m_settings.getTargetColumn() == null) {
        List<DataColumnSpec> possibleTargets = new ArrayList<DataColumnSpec>();
        for (DataColumnSpec colSpec : inSpec) {
            if (colSpec.getType().isCompatible(DoubleValue.class)) {
                possibleTargets.add(colSpec);
            }
        }
        if (possibleTargets.size() > 1) {
            String colName = possibleTargets.get(possibleTargets.size() - 1).getName();
            m_settings.setTargetColumn(colName);
            String warning = "The target column is not set. Using " + colName;
            m_warningMessage = (m_warningMessage == null ? "" : m_warningMessage + "\n") + warning;
        } else if (possibleTargets.size() == 1) {
            m_settings.setTargetColumn(possibleTargets.get(0).getName());
        } else {
            throw new InvalidSettingsException("No column in " + "spec with numeric data.");
        }
    }
    FilterResult colFilter = m_settings.getFilterConfiguration().applyTo(inSpec);
    List<String> inputCols = new ArrayList<String>();
    inputCols.addAll(Arrays.asList(colFilter.getIncludes()));
    inputCols.remove(m_settings.getTargetColumn());
    // remove all columns that should not be used
    inputCols.removeAll(exclude);
    if (inputCols.isEmpty()) {
        throw new InvalidSettingsException("At least one column must be included.");
    }
    DataColumnSpec targetColSpec = null;
    List<DataColumnSpec> regressorColSpecs = new ArrayList<DataColumnSpec>();
    // Check type of target and input columns
    for (int i = 0; i < inSpec.getNumColumns(); i++) {
        DataColumnSpec colSpec = inSpec.getColumnSpec(i);
        String colName = colSpec.getName();
        if (m_settings.getTargetColumn().equals(colName)) {
            if (colSpec.getType().isCompatible(DoubleValue.class)) {
                targetColSpec = colSpec;
            } else {
                throw new InvalidSettingsException("Type of column \"" + colName + "\" is not numeric.");
            }
        } else if (inputCols.contains(colName)) {
            if (colSpec.getType().isCompatible(DoubleValue.class) || colSpec.getType().isCompatible(NominalValue.class)) {
                regressorColSpecs.add(colSpec);
            } else {
                throw new InvalidSettingsException("Type of column \"" + colName + "\" is not one of the allowed types, " + "which are numeric or nominal.");
            }
        }
    }
    if (targetColSpec != null) {
        String[] learnerCols = new String[regressorColSpecs.size() + 1];
        for (int i = 0; i < regressorColSpecs.size(); i++) {
            learnerCols[i] = regressorColSpecs.get(i).getName();
        }
        learnerCols[learnerCols.length - 1] = targetColSpec.getName();
        PMMLPortObjectSpecCreator creator = new PMMLPortObjectSpecCreator(pmmlSpec, inSpec);
        creator.setTargetCols(Arrays.asList(targetColSpec));
        creator.setLearningCols(regressorColSpecs);
        m_pmmlOutSpec = creator.createSpec();
        m_learner = new Learner(m_pmmlOutSpec, m_settings.getIncludeConstant(), m_settings.getOffsetValue(), m_settings.getMissingValueHandling2().equals(MissingValueHandling.fail));
    } else {
        throw new InvalidSettingsException("The target is " + "not in the input.");
    }
}

Example 17 with PMMLPortObjectSpecCreator

use of org.knime.core.node.port.pmml.PMMLPortObjectSpecCreator in project knime-core by knime.

the class LinReg2LearnerNodeModel method execute.

/**
 * {@inheritDoc}
 */
@Override
protected PortObject[] execute(final PortObject[] inObjects, final ExecutionContext exec) throws Exception {
    final BufferedDataTable data = (BufferedDataTable) inObjects[0];
    // cache the entire table as otherwise the color information
    // may be lost (filtering out the "colored" column)
    m_rowContainer = new DefaultDataArray(data, m_settings.getScatterPlotFirstRow(), m_settings.getScatterPlotRowCount());
    DataTableSpec tableSpec = data.getDataTableSpec();
    // handle the optional PMML input
    PMMLPortObject inPMMLPort = m_pmmlInEnabled ? (PMMLPortObject) inObjects[1] : null;
    PMMLPortObjectSpec inPMMLSpec = null;
    if (inPMMLPort != null) {
        inPMMLSpec = inPMMLPort.getSpec();
    } else {
        PMMLPortObjectSpecCreator creator = new PMMLPortObjectSpecCreator(tableSpec);
        inPMMLSpec = creator.createSpec();
        inPMMLPort = new PMMLPortObject(inPMMLSpec);
    }
    LinReg2Learner learner = new LinReg2Learner(new PortObjectSpec[] { tableSpec, inPMMLSpec }, m_pmmlInEnabled, m_settings);
    m_content = learner.execute(new PortObject[] { data, inPMMLPort }, exec);
    if (learner.getWarningMessage() != null && learner.getWarningMessage().length() > 0) {
        setWarningMessage(learner.getWarningMessage());
    }
    // third argument is ignored since we provide a port
    PMMLPortObject outPMMLPort = new PMMLPortObject((PMMLPortObjectSpec) learner.getOutputSpec()[0], inPMMLPort, null);
    PMMLGeneralRegressionTranslator trans = new PMMLGeneralRegressionTranslator(m_content.createGeneralRegressionContent());
    outPMMLPort.addModelTranslater(trans);
    final String warningMessage = m_content.getWarningMessage();
    if (warningMessage != null) {
        setWarningMessage(getWarningMessage() == null ? warningMessage : (getWarningMessage() + "\n" + warningMessage));
    }
    return new PortObject[] { outPMMLPort, m_content.createTablePortObject(exec) };
}

Example 18 with PMMLPortObjectSpecCreator

use of org.knime.core.node.port.pmml.PMMLPortObjectSpecCreator in project knime-core by knime.

the class PolyRegLearnerNodeModel method createModelSpec.

private PMMLPortObjectSpec createModelSpec(final PMMLPortObjectSpec inModelSpec, final DataTableSpec inDataSpec) throws InvalidSettingsException {
    String[] selectedCols = computeSelectedColumns(inDataSpec);
    DataColumnSpec[] usedColumns = new DataColumnSpec[selectedCols.length + 1];
    int k = 0;
    Set<String> hash = new HashSet<String>(Arrays.asList(selectedCols));
    for (DataColumnSpec dcs : inDataSpec) {
        if (hash.contains(dcs.getName())) {
            usedColumns[k++] = dcs;
        }
    }
    usedColumns[k++] = inDataSpec.getColumnSpec(m_settings.getTargetColumn());
    DataTableSpec tableSpec = new DataTableSpec(usedColumns);
    PMMLPortObjectSpecCreator crea = new PMMLPortObjectSpecCreator(inModelSpec, inDataSpec);
    crea.setLearningCols(tableSpec);
    crea.setTargetCol(usedColumns[k - 1]);
    return crea.createSpec();
}

Example 19 with PMMLPortObjectSpecCreator

use of org.knime.core.node.port.pmml.PMMLPortObjectSpecCreator in project knime-core by knime.

the class ClusterNodeModel method configure.

/**
 * Returns <code>true</code> always and passes the current input spec to
 * the output spec which is identical to the input specification - after
 * all, we are building cluster centers in the original feature space.
 *
 * @param inSpecs the specifications of the input port(s) - should be one
 * @return the copied input spec
 * @throws InvalidSettingsException if PMML incompatible type was found
 */
@Override
protected PortObjectSpec[] configure(final PortObjectSpec[] inSpecs) throws InvalidSettingsException {
    DataTableSpec spec = (DataTableSpec) inSpecs[0];
    // input is output spec with all double compatible values set to
    // Double.
    m_dimension = spec.getNumColumns();
    // Find out which columns we can use (must be Double compatible)
    // Note that, for simplicity, we still use the entire dimensionality
    // for cluster prototypes below and simply ignore useless columns.
    m_ignoreColumn = new boolean[m_dimension];
    m_nrIgnoredColumns = 0;
    LinkedList<String> includes = new LinkedList<String>();
    includes.addAll(m_usedColumns.getIncludeList());
    LinkedList<String> excludes = new LinkedList<String>();
    excludes.addAll(m_usedColumns.getExcludeList());
    LinkedList<String> includes2 = new LinkedList<String>();
    includes2.addAll(m_usedColumns.getIncludeList());
    LinkedList<String> excludes2 = new LinkedList<String>();
    excludes2.addAll(m_usedColumns.getExcludeList());
    // First check if all incoming columns are either excluded or included
    for (String col : spec.getColumnNames()) {
        if (m_usedColumns.getIncludeList().contains(col)) {
            includes2.remove(col);
        } else if (m_usedColumns.getExcludeList().contains(col)) {
            excludes2.remove(col);
        } else {
            includes.add(col);
        }
    }
    // Leftover included columns that do not exist in the incoming table
    for (String col : includes2) {
        includes.remove(col);
    }
    // Same for excluded columns
    for (String col : excludes2) {
        excludes.remove(col);
    }
    m_usedColumns.setExcludeList(excludes);
    m_usedColumns.setIncludeList(includes);
    if (m_usedColumns.isKeepAllSelected()) {
        boolean hasNumericColumn = false;
        for (DataColumnSpec colSpec : spec) {
            if (colSpec.getType().isCompatible(DoubleValue.class)) {
                hasNumericColumn = true;
                break;
            }
        }
        if (!hasNumericColumn) {
            throw new InvalidSettingsException("No numeric columns in input");
        }
    } else {
        // double compatible columns
        if (m_usedColumns.getIncludeList().size() == 0 && m_usedColumns.getExcludeList().size() == 0) {
            List<String> includedColumns = new ArrayList<String>();
            List<String> excludedColumns = new ArrayList<String>();
            for (int i = 0; i < spec.getNumColumns(); i++) {
                DataColumnSpec colSpec = spec.getColumnSpec(i);
                if (colSpec.getType().isCompatible(DoubleValue.class)) {
                    includedColumns.add(colSpec.getName());
                } else {
                    excludedColumns.add(colSpec.getName());
                }
            }
            // set all double compatible columns as include list
            m_usedColumns.setIncludeList(includedColumns);
            m_usedColumns.setExcludeList(excludedColumns);
        }
        // check if some columns are included
        if (m_usedColumns.getIncludeList().size() <= 0) {
            setWarningMessage("No column in include list! Produces one huge cluster");
        }
    }
    addExcludeColumnsToIgnoreList(spec);
    DataTableSpec appendedSpec = createAppendedSpec(spec);
    // return spec for data and model outport!
    PMMLPortObjectSpec pmmlSpec;
    if (m_pmmlInEnabled) {
        pmmlSpec = (PMMLPortObjectSpec) inSpecs[1];
    } else {
        pmmlSpec = new PMMLPortObjectSpecCreator(spec).createSpec();
    }
    if (m_outputCenters) {
        return new PortObjectSpec[] { appendedSpec, createClusterCentersSpec(spec), createPMMLSpec(pmmlSpec, spec) };
    } else {
        return new PortObjectSpec[] { appendedSpec, createPMMLSpec(pmmlSpec, spec) };
    }
}

Example 20 with PMMLPortObjectSpecCreator

use of org.knime.core.node.port.pmml.PMMLPortObjectSpecCreator in project knime-core by knime.

the class FuzzyClusterNodeModel method createPMMLPortObjectSpec.

private PMMLPortObjectSpec createPMMLPortObjectSpec(final PMMLPortObjectSpec pmmlSpec, final DataTableSpec inspec, final List<String> learningCols) {
    PMMLPortObjectSpecCreator pmmlSpecCreator = new PMMLPortObjectSpecCreator(pmmlSpec, inspec);
    pmmlSpecCreator.setLearningColsNames(learningCols);
    return pmmlSpecCreator.createSpec();
}