Java tutorial
/******************************************************************************* * This file is part of OpenNMS(R). * * Copyright (C) 2010-2015 The OpenNMS Group, Inc. * OpenNMS(R) is Copyright (C) 1999-2015 The OpenNMS Group, Inc. * * OpenNMS(R) is a registered trademark of The OpenNMS Group, Inc. * * OpenNMS(R) is free software: you can redistribute it and/or modify * it under the terms of the GNU Affero General Public License as published * by the Free Software Foundation, either version 3 of the License, * or (at your option) any later version. * * OpenNMS(R) is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Affero General Public License for more details. * * You should have received a copy of the GNU Affero General Public License * along with OpenNMS(R). If not, see: * http://www.gnu.org/licenses/ * * For more information contact: * OpenNMS(R) Licensing <license@opennms.org> * http://www.opennms.org/ * http://www.opennms.com/ *******************************************************************************/ package org.opennms.netmgt.measurements.filters.impl; import java.util.Date; import java.util.Map; import org.opennms.netmgt.integrations.R.RScriptException; import org.opennms.netmgt.integrations.R.RScriptExecutor; import org.opennms.netmgt.integrations.R.RScriptInput; import org.opennms.netmgt.integrations.R.RScriptOutput; import org.opennms.netmgt.measurements.api.Filter; import org.opennms.netmgt.measurements.api.FilterInfo; import org.opennms.netmgt.measurements.api.FilterParam; import org.opennms.netmgt.measurements.filters.impl.Utils.TableLimits; import org.slf4j.Logger; import org.slf4j.LoggerFactory; import com.google.common.base.Preconditions; import com.google.common.collect.ImmutableTable; import com.google.common.collect.Maps; import com.google.common.collect.RowSortedTable; import com.google.common.collect.TreeBasedTable; /** * Performs Holt-Winters forecasting on a given column of * the data source with R. * * @author jwhite */ @FilterInfo(name = "HoltWinters", description = "Performs Holt-Winters forecasting.", backend = "R") public class HWForecast implements Filter { private static final Logger LOG = LoggerFactory.getLogger(HWForecast.class); private static final String PATH_TO_R_SCRIPT = "/org/opennms/netmgt/measurements/filters/impl/holtWinters.R"; @FilterParam(key = "inputColumn", required = true, displayName = "Input", description = "Input column.") private String m_inputColumn; @FilterParam(key = "outputPrefix", value = "HW", displayName = "Output", description = "Output prefix.") private String m_outputPrefix; @FilterParam(key = "numPeriodsToForecast", value = "3", displayName = "# Periods", description = "Number of periods to forecast.") private int m_numPeriodsToForecast; @FilterParam(key = "periodInSeconds", required = true, displayName = "Period", description = "Size of a period in seconds.") private long m_periodInSeconds; @FilterParam(key = "confidenceLevel", value = "0.95", displayName = "Level", description = "Probability used for confidence bounds. Set this to 0 in order to disable the bounds.") private double m_confidenceLevel; protected HWForecast() { } public HWForecast(String outputPrefix, String inputColumn, int numPeriodsToForecast, long periodInSeconds, double confidenceLevel) { m_outputPrefix = outputPrefix; m_inputColumn = inputColumn; m_numPeriodsToForecast = numPeriodsToForecast; m_periodInSeconds = periodInSeconds; m_confidenceLevel = confidenceLevel; } @Override public void filter(RowSortedTable<Long, String, Double> table) throws RScriptException { Preconditions.checkArgument(table.containsColumn(TIMESTAMP_COLUMN_NAME), String.format("Data source must have a '%s' column.", Filter.TIMESTAMP_COLUMN_NAME)); // Determine the index of the first and last non-NaN values // Assume the values between these are contiguous TableLimits limits = Utils.getRowsWithValues(table, m_inputColumn); // Make sure we have some samples long numSampleRows = limits.lastRowWithValues - limits.firstRowWithValues; if (numSampleRows < 1) { LOG.error( "Insufficient values in column for forecasting. Excluding forecast columns from data source."); return; } // Determine the step size Date lastTimestamp = new Date(table.get(limits.lastRowWithValues, TIMESTAMP_COLUMN_NAME).longValue()); long stepInMs = (long) (table.get(limits.lastRowWithValues, TIMESTAMP_COLUMN_NAME) - table.get(limits.lastRowWithValues - 1, Filter.TIMESTAMP_COLUMN_NAME)); // Calculate the number of samples per period int numSamplesPerPeriod = (int) Math.floor(m_periodInSeconds * 1000 / stepInMs); numSamplesPerPeriod = Math.max(1, numSamplesPerPeriod); // Calculate the number of steps to forecast int numForecasts = numSamplesPerPeriod * m_numPeriodsToForecast; // Script arguments Map<String, Object> arguments = Maps.newHashMap(); arguments.put("columnToForecast", m_inputColumn); arguments.put("numSamplesPerSeason", numSamplesPerPeriod); arguments.put("numForecasts", numForecasts); arguments.put("confidenceLevel", m_confidenceLevel); // Array indices in R start at 1 arguments.put("firstIndex", limits.firstRowWithValues + 1); arguments.put("lastIndex", limits.lastRowWithValues + 1); // Make the forecasts RScriptExecutor executor = new RScriptExecutor(); RScriptOutput output = executor.exec(PATH_TO_R_SCRIPT, new RScriptInput(table, arguments)); ImmutableTable<Long, String, Double> outputTable = output.getTable(); // The output table contains the fitted values, followed // by the requested number of forecasted values int numOutputRows = outputTable.rowKeySet().size(); int numFittedValues = numOutputRows - numForecasts; // Add the fitted values to rows where the input column has values for (long i = 0; i < numFittedValues; i++) { long idxTarget = i + (numSampleRows - numFittedValues) + limits.firstRowWithValues + 1; table.put(idxTarget, m_outputPrefix + "Fit", outputTable.get(i, "fit")); } // Append the forecasted values and include the time stamp with the appropriate step for (long i = numFittedValues; i < numOutputRows; i++) { long idxForecast = i - numFittedValues + 1; long idxTarget = limits.lastRowWithValues + idxForecast; if (m_confidenceLevel > 0) { table.put(idxTarget, m_outputPrefix + "Fit", outputTable.get(i, "fit")); table.put(idxTarget, m_outputPrefix + "Lwr", outputTable.get(i, "lwr")); table.put(idxTarget, m_outputPrefix + "Upr", outputTable.get(i, "upr")); } table.put(idxTarget, TIMESTAMP_COLUMN_NAME, (double) new Date(lastTimestamp.getTime() + stepInMs * idxForecast).getTime()); } } public static void checkForecastSupport() throws RScriptException { // Verify the HW filter HWForecast forecastFilter = new HWForecast("HW", "X", 1, 1, 0.95); // Use constant values for the Y column RowSortedTable<Long, String, Double> table = TreeBasedTable.create(); for (long i = 0; i < 100; i++) { table.put(i, Filter.TIMESTAMP_COLUMN_NAME, (double) (i * 1000)); table.put(i, "X", 1.0d); } // Apply the filter forecastFilter.filter(table); } }