/* * To change this template, choose Tools | Templates * and open the template in the editor. */ package org.virbo.idlsupport; import java.text.NumberFormat; import java.util.ArrayList; import java.util.LinkedHashMap; import java.util.List; import java.util.Map; import java.util.logging.Level; import java.util.logging.Logger; import org.das2.qds.buffer.FloatDataSet; import org.das2.qds.buffer.LongDataSet; import org.das2.datum.LoggerManager; import org.das2.datum.Units; import org.das2.datum.UnitsConverter; import org.das2.util.monitor.NullProgressMonitor; import org.das2.util.monitor.ProgressMonitor; import org.das2.qds.DataSetOps; import org.das2.qds.DataSetUtil; import org.das2.qds.FDataSet; import org.das2.qds.QDataSet; import org.das2.qds.QubeDataSetIterator; import org.das2.qds.SemanticOps; /** * See http://autoplot.org/IDL and http://autoplot.org/Matlab which show how this is used in the environments. * @author jbf */ public abstract class QDataSetBridge { private static final Logger logger= LoggerManager.getLogger("qdataset.bridge"); QDataSet ds; Exception exception; String name; Map datasets; /** * this contains the names of things that should resolve when slicing. */ Map sliceDep; Map names; List prefUnits; // convert to these if possible double fill; // use this fill value float ffill; long lfill= -999999999999999999L; // hope this isn't used. boolean useFill=false; // true means convert fill values boolean debug= false; QDataSetBridge() { datasets = new LinkedHashMap<>(); names = new LinkedHashMap<>(); sliceDep= new LinkedHashMap<>(); prefUnits= new ArrayList(); } /** * set the preferred units for the data. The code will convert * to these units when a converter is found. * If a preference is indicated when it is convertible to an existing * preference, the existing preference is removed. * See clearPreferredUnits to remove all preferences. * Example units strings (capitalization matters): **
{@code
     *    seconds since 2010-01-01T00:00
     *    days since 2010-01-01T00:00
     *    Hz
     *    kHz
     *    MHz
     *}
* @param sunit */ public void setPreferredUnits( String sunit ) { Units unit; if ( sunit.contains(" since ") ) { unit= Units.lookupUnits(sunit); } else { unit= Units.getByName(sunit); } boolean add= true; for ( int i=0; i apds.setFillValue, 89. * % setFillValue(float) * IDL> apds.setFillValue, 89.d * % setFillValue(double) * @param d */ public void setFillValue( double d ) { setFillDouble( d ); } public void setFillValue( float f ) { this.ffill= f; this.fill= (double)f; // danger this.useFill= true; } /** * set the value to return when the data is invalid, when the data is known to be stored as doubles * (8-byte numbers). Python JPype doesn't allow for operator overloading, so this should be used. * @param d */ public void setFillDouble( double d ) { this.fill= d; this.ffill= (float)d; // danger this.useFill= true; } protected String filter = ""; public static final String PROP_FILTER = "filter"; /** * get the filter that is applied to the data immediately after it is loaded. * @return the filter string, empty string means no filters. */ public String getFilter() { return filter; } /** * set the filter that is applied to the data immediately after it is loaded. * @param filter filter string like "|histogram()" or "" for no filters. */ public void setFilter(String filter) { this.filter = filter; } /** * turn on/off debug messages * @param debug */ public void setDebug( boolean debug ) { System.err.println("setting debug="+debug); this.debug= debug; } /** * don't use fill value. */ public void clearFillValue() { this.useFill= false; } /** * performs the read. Note no progress status is available and this blocks until the read is done. * doGetDataSet(mon) should be called if progress is needed. * 2011-01-01: getStatus or getStatusMessage should be called afterwards to check the result of the load, this will no longer throw an exception. */ public void doGetDataSet() { this.exception= null; try { this.ds = getDataSet( new NullProgressMonitor() ); if ( this.filter.length()>0 ) { ds= DataSetOps.sprocess( this.filter, ds, new NullProgressMonitor() ); } datasets.clear(); name = nameFor(ds); datasets.put(name, ds); if ( SemanticOps.isBundle(ds) ) { String[] ss= DataSetOps.bundleNames(ds); for ( String s: ss ) { datasets.put( s, DataSetOps.unbundle(ds,s) ); } } for (int i = 0; i < ds.rank(); i++) { QDataSet dep = (QDataSet) ds.property("DEPEND_" + i); if (dep != null) datasets.put(nameFor(dep), dep); QDataSet depslice= (QDataSet) ds.property("DEPEND_" + i, 0 ); if ( depslice!=null ) { sliceDep.put( nameFor(depslice,true), "DEPEND_"+i ); } } QDataSet ads; int i; i=0; while ( (ads=(QDataSet)ds.property( "PLANE_"+i ))!=null ) { datasets.put( nameFor( ads ), ads ); i++; } i=0; while ( (ads=(QDataSet)ds.property( "PLANE_"+i, 0 ))!=null ) { sliceDep.put( nameFor( ads,true ), "PLANE_"+i ); i++; } } catch ( Exception ex ) { this.exception= ex; logger.log(Level.WARNING, null, ex ); } } /** * This initiates the read on a separate thread, so this does not * block and should be used with caution. See getProgressMonitor for * use. Note this is more advanced and is intended to support use in * other software (e.g. PaPCo). * * Note because there is one exception that is stored, a QDataSetBridge object * is only able to load one dataset at a time. Simultaneous loads should * be done with multiple QDataSetBridge objects. * * @param mon progress monitor from getProgressMonitor() */ public void doGetDataSet(final ProgressMonitor mon) { this.exception= null; Runnable run = new Runnable() { @Override public void run() { datasets.clear(); name= ""; try { ds = getDataSet(mon); if ( ds==null ) return; if ( filter.length()>0 ) { ds= DataSetOps.sprocess( filter, ds, new NullProgressMonitor() ); } } catch (Exception ex) { exception= ex; logger.log( Level.WARNING, null, ex ); mon.setProgressMessage("EXCEPTION"); mon.finished(); return; } name = nameFor(ds); datasets.put(name, ds); for (int i = 0; i < ds.rank(); i++) { QDataSet dep = (QDataSet) ds.property("DEPEND_" + i); if (dep != null) datasets.put(nameFor(dep), dep); QDataSet depslice= (QDataSet) ds.property("DEPEND_" + i, 0 ); if ( depslice!=null ) { sliceDep.put( nameFor(depslice), "DEPEND_"+i ); } } } }; new Thread(run).start(); } /** * return the Exception from the last doGetDataSet call. * @return */ public Exception getException() { return exception; } /** * returns 0 for last get operation successful * @return */ public int getStatus() { return exception==null ? 0 : 1; } /** * returns "" for last operation successful, or non-empty indicating the problem. Note * getException will return the Java exception for deeper inspection. * @return "" or the error message */ public String getStatusMessage() { if ( exception==null ) { return ""; } else { String s= exception.getMessage(); if ( s!=null && s.length()>0 ) { return s; } else { return exception.toString(); // sorry! } } } /** * return the name used to refer to the dataset. * This may add to the list of datasets. * @param dep0 the dataset * @return the name for the dataset. */ public synchronized String nameFor(QDataSet dep0) { return nameFor( dep0, false ); } /** * return the name used to refer to the dataset. * This may add to the list of datasets. * @param dep0 the dataset * @return the name for the dataset. */ private synchronized String nameFor(QDataSet dep0, boolean onlySlice ) { String name1 = names.get(dep0); if ( name1==null ) { name1 = (String) dep0.property(QDataSet.NAME); if ( name1 == null || datasets.containsKey(name1) ) { name1 = "ds_" + names.size(); } names.put(dep0, name1); if ( datasets.containsKey(name1) ) { throw new IllegalArgumentException("dataset name is already taken: "+name1 ); } if ( !onlySlice ) datasets.put( name1, dep0 ); } return name1; } /** * implementations should provide this method for making the data accessible. * @param mon * @return * @throws Exception if any problem occurs with the read, it will be available to clients in getException or getStatusMessage */ abstract QDataSet getDataSet(ProgressMonitor mon) throws Exception; /** * returns an object that can be used to monitor the progress of a download. * * mon= qds->getProgressMonitor(); * qds->doGetDataSet( mon ) * while ( ! mon->isFinished() ) do begin * print, strtrim( mon->getTaskProgress(), 2 ) + " " + strtrim( mon->getTaskSize(), 2 ) * wait, 0.2 ; don't overload the thread * endwhile * * @return */ public ProgressMonitor getProgressMonitor() { return new NullProgressMonitor(); } public void values(String name, double[] result) { if ( debug ) { System.err.println("reading "+name+" into double["+result.length+"]" ); } QDataSet ds1 = datasets.get(name); copyValues( ds1, result ); } public void values(String name, double[][] result) { if ( debug ) { System.err.println("reading "+name+" into double["+result.length+","+result[0].length+"]" ); } QDataSet ds1 = datasets.get(name); copyValues( ds1, result ); } public void values(String name, double[][][] result) { if ( debug ) { System.err.println("reading "+name+" into double["+result.length +","+result[0].length +","+result[0][0].length+"]" ); } QDataSet ds1 = datasets.get(name); copyValues( ds1, result ); } public void values(String name, double[][][][] result) { if ( debug ) { System.err.println("reading "+name+" into double["+result.length +","+result[0].length +","+result[0][0].length +","+result[0][0][0].length+"]" ); } QDataSet ds1 = datasets.get(name); copyValues( ds1, result ); } public void values(double[] result) { values(this.name(), result); } public void values(double[][] result) { values(this.name(), result); } public void values(double[][][] result) { values(this.name(), result); } public void values(double[][][][] result) { values(this.name(), result); } /** * returns the converter if there is one. If no converter is * registered, then UnitsConverter.IDENTITY is returned. * @param ds1 * @return */ private UnitsConverter maybeGetConverter( QDataSet ds1 ) { Units u= SemanticOps.getUnits(ds1); UnitsConverter uc= UnitsConverter.IDENTITY; if ( prefUnits!=null ) { for ( Units prefUnit: prefUnits ) { if ( prefUnit.isConvertableTo(u) ) { uc= u.getConverter(prefUnit); if ( uc!=UnitsConverter.IDENTITY ) { if ( debug ) { System.err.println("Using units converter to get "+prefUnit ); } } } } } return uc; } /* -- convert qubes to float arrays -- */ private void copyValues( QDataSet ds1, long[] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); if ( debug ) { System.err.println("copyValues rank1 into float using "+uc); } for (int i0 = 0; i0 < ds1.length(); i0++) { if ( useFill && wds.value(i0)==0 ) { result[i0] = lfill; } else { result[i0] = (long)uc.convert( ds1.value(i0) ); } } } private void copyValues( QDataSet ds1, long[][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { if ( useFill && wds.value(i0, i1 )==0 ) { result[i0][i1] = lfill; } else { result[i0][i1] = (long)uc.convert( ds1.value(i0, i1) ); } } } } private void copyValues( QDataSet ds1, long[][][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { for (int i2 = 0; i2 < ds1.length(i0,i1); i2++) { if ( useFill && wds.value(i0, i1, i2 )==0 ) { result[i0][i1][i2] = lfill; } else { result[i0][i1][i2] = (long)uc.convert( ds1.value(i0, i1, i2) ); } } } } } private void copyValues( QDataSet ds1, long[][][][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { for (int i2 = 0; i2 < ds1.length(i0,i1); i2++) { for (int i3 = 0; i3 < ds1.length(i0,i1,i2); i2++) { if ( useFill && wds.value(i0, i1, i2, i3 )==0 ) { result[i0][i1][i2][i3] = lfill; } else { result[i0][i1][i2][i3] = (long)uc.convert( ds1.value(i0, i1, i2, i3 ) ); } } } } } } /* -- convert qubes to float arrays -- */ private void copyValues( QDataSet ds1, float[] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); if ( debug ) { System.err.println("copyValues rank1 into float using "+uc); } for (int i0 = 0; i0 < ds1.length(); i0++) { if ( useFill && wds.value(i0)==0 ) { result[i0] = ffill; } else { result[i0] = (float)uc.convert( ds1.value(i0) ); } } } private void copyValues( QDataSet ds1, float[][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { if ( useFill && wds.value(i0, i1 )==0 ) { result[i0][i1] = ffill; } else { result[i0][i1] = (float)uc.convert( ds1.value(i0, i1) ); } } } } private void copyValues( QDataSet ds1, float[][][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { for (int i2 = 0; i2 < ds1.length(i0,i1); i2++) { if ( useFill && wds.value(i0, i1, i2 )==0 ) { result[i0][i1][i2] = ffill; } else { result[i0][i1][i2] = (float)uc.convert( ds1.value(i0, i1, i2) ); } } } } } private void copyValues( QDataSet ds1, float[][][][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { for (int i2 = 0; i2 < ds1.length(i0,i1); i2++) { for (int i3 = 0; i3 < ds1.length(i0,i1,i2); i2++) { if ( useFill && wds.value(i0, i1, i2, i3 )==0 ) { result[i0][i1][i2][i3] = ffill; } else { result[i0][i1][i2][i3] = (float)uc.convert( ds1.value(i0, i1, i2, i3 ) ); } } } } } } /* -- convert qubes to double arrays -- */ private void copyValues( QDataSet ds1, double[] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); if ( debug ) { System.err.println("copyValues rank1 into double using "+uc); } for (int i0 = 0; i0 < ds1.length(); i0++) { if ( useFill && wds.value(i0)==0 ) { result[i0] = fill; } else { result[i0] = uc.convert( ds1.value(i0) ); } } } private void copyValues( QDataSet ds1, double[][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { if ( useFill && wds.value(i0, i1 )==0 ) { result[i0][i1] = fill; } else { result[i0][i1] = uc.convert( ds1.value(i0, i1) ); } } } } private void copyValues( QDataSet ds1, double[][][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { for (int i2 = 0; i2 < ds1.length(i0,i1); i2++) { if ( useFill && wds.value(i0, i1, i2 )==0 ) { result[i0][i1][i2] = fill; } else { result[i0][i1][i2] = uc.convert( ds1.value(i0, i1, i2) ); } } } } } private void copyValues( QDataSet ds1, double[][][][] result ) { UnitsConverter uc= maybeGetConverter(ds1); QDataSet wds= DataSetUtil.weightsDataSet(ds1); for (int i0 = 0; i0 < ds1.length(); i0++) { for (int i1 = 0; i1 < ds1.length(i0); i1++) { for (int i2 = 0; i2 < ds1.length(i0,i1); i2++) { for (int i3 = 0; i3 < ds1.length(i0,i1,i2); i2++) { if ( useFill && wds.value(i0, i1, i2, i3 )==0 ) { result[i0][i1][i2][i3] = fill; } else { result[i0][i1][i2][i3] = uc.convert( ds1.value(i0, i1, i2, i3 ) ); } } } } } } /** * accessor for non-qube * @param name * @param i * @param result */ public void slice(String name, int i, double[] result) { if ( debug ) { System.err.println("reading "+name+"["+i+"] into double["+result.length +"]" ); } QDataSet ds1; if ( datasets.get(name)!=null ) { ds1 = datasets.get(name).slice(i); } else { throw new IllegalArgumentException("did not find dataset name="+name ); } copyValues( ds1, result ); } /** * accessor for non-qube * @param name * @param i * @param result */ public void slice(String name, int i, double[][] result) { if ( debug ) { System.err.println("reading "+name+"["+i+"] into double["+result.length +","+result[0].length +"]" ); } QDataSet ds1; if ( datasets.get(name)!=null ) { ds1 = datasets.get(name).slice(i); } else { throw new IllegalArgumentException("did not find dataset name="+name ); } copyValues( ds1, result ); } public void slice(String name, int i, double[][][] result) { if ( debug ) { System.err.println("reading "+name+"["+i+"] into double["+result.length +","+result[0].length +","+result[0][0].length +"]" ); } QDataSet ds1; if ( datasets.get(name)!=null ) { ds1 = datasets.get(name).slice(i); } else { throw new IllegalArgumentException("did not find dataset name="+name ); } copyValues( ds1, result ); } public void slice(int i, double[] result) { slice(this.name(), i, result); } public void slice(int i, double[][] result) { slice(this.name(), i, result); } public void slice(int i, double[][][] result) { slice(this.name(), i, result); } /** * return an 1,2,or 3-D array of doubles or floats containing the values * in the specified dataset. * @param name * @return */ public Object values(String name) { if ( debug ) { System.err.println("reading values for dataset " + name ); } QDataSet ds1 = datasets.get(name); if ( ds1==null ) { throw new IllegalArgumentException("No datasets with the name: "+name); } UnitsConverter uc= maybeGetConverter(ds1); if ( ds1 instanceof FDataSet || ds1 instanceof FloatDataSet ) { switch (ds1.rank()) { case 1: { float[] result = new float[ds1.length()]; copyValues(ds1, result); return result; } case 2: { float[][] result = new float[ds1.length()][ds1.length(0)]; copyValues(ds1, result); return result; } case 3: { float[][][] result = new float[ds1.length()][ds1.length(0)][ds1.length(0, 0)]; copyValues(ds1, result); return result; } case 4: { float[][][][] result = new float[ds1.length()][ds1.length(0)][ds1.length(0, 0)][ds1.length(0,0,0)]; copyValues(ds1, result); return result; } default: throw new IllegalArgumentException("rank limit"); } } else if ( ds1 instanceof LongDataSet && uc==UnitsConverter.IDENTITY ) { // Special support for CDF TT2000 switch (ds1.rank()) { case 1: { long[] result = new long[ds1.length()]; copyValues(ds1, result); return result; } case 2: { long[][] result = new long[ds1.length()][ds1.length(0)]; copyValues(ds1, result); return result; } case 3: { long[][][] result = new long[ds1.length()][ds1.length(0)][ds1.length(0, 0)]; copyValues(ds1, result); return result; } case 4: { long[][][][] result = new long[ds1.length()][ds1.length(0)][ds1.length(0, 0)][ds1.length(0,0,0)]; copyValues(ds1, result); return result; } default: throw new IllegalArgumentException("rank limit"); } } else { switch (ds1.rank()) { case 1: { double[] result = new double[ds1.length()]; values(name, result); return result; } case 2: { double[][] result = new double[ds1.length()][ds1.length(0)]; values(name, result); return result; } case 3: { double[][][] result = new double[ds1.length()][ds1.length(0)][ds1.length(0, 0)]; values(name, result); return result; } case 4: { double[][][][] result = new double[ds1.length()][ds1.length(0)][ds1.length(0, 0)][ds1.length(0,0,0)]; values(name, result); return result; } default: throw new IllegalArgumentException("rank limit"); } } } /** * return the i-th 1 or 2-D array of doubles or floats containing the values * in the specified rank 2 or 3 dataset. This is to support non-qube datasets. * @param name * @param i * @return */ public Object slice( String name, int i ) { if ( debug ) { System.err.println("reading values for slice " + i + " of dataset " + name ); } QDataSet ds1 = datasets.get(name); if ( ds1==null && sliceDep.containsKey(name) ) { return sliceDep(name,i); } if ( ds1==null ) { throw new IllegalArgumentException("No such dataset: "+name ); } switch (ds1.rank()) { case 1: throw new IllegalArgumentException("dataset is rank 1, slice not allowed"); case 2: { double[] result = new double[ds1.length(i)]; slice(name, i, result); return result; } case 3: { double[][] result = new double[ds1.length(i)][ds1.length(i,0)]; slice(name, i,result); return result; } case 4: { double[][][] result = new double[ds1.length(i)][ds1.length(i,0)][ds1.length(i,0,0)]; slice(name, i,result); return result; } default: throw new IllegalArgumentException("rank limit"); } } /** * we have to slice the main dataset first, then copy the values of the depend dataset * @param name * @param i * @return */ private Object sliceDep( String name, int i ) { QDataSet ds1= (QDataSet) datasets.get(this.name).slice(i).property(sliceDep.get(name)); UnitsConverter uc= maybeGetConverter(ds1); if ( ds1 instanceof FDataSet || ds1 instanceof FloatDataSet) { //TODO: don't forget about BufferDataSet FloatDataSet switch (ds1.rank()) { case 1: { float[] result = new float[ds1.length()]; copyValues( ds1, result ); return result; } case 2: { float[][] result = new float[ds1.length()][ds1.length(0)]; copyValues( ds1, result ); return result; } case 3: { float[][][] result = new float[ds1.length()][ds1.length(0)][ds1.length(0,0)]; copyValues( ds1, result ); return result; } default: throw new IllegalArgumentException("rank limit"); } } else if ( ds1 instanceof LongDataSet && uc==UnitsConverter.IDENTITY ) { // Special support for CDF TT2000 switch (ds1.rank()) { case 1: { long[] result = new long[ds1.length()]; copyValues(ds1, result); return result; } case 2: { long[][] result = new long[ds1.length()][ds1.length(0)]; copyValues(ds1, result); return result; } case 3: { long[][][] result = new long[ds1.length()][ds1.length(0)][ds1.length(0, 0)]; copyValues(ds1, result); return result; } case 4: { long[][][][] result = new long[ds1.length()][ds1.length(0)][ds1.length(0, 0)][ds1.length(0,0,0)]; copyValues(ds1, result); return result; } default: throw new IllegalArgumentException("rank limit"); } } else { switch (ds1.rank()) { case 1: { double[] result = new double[ds1.length()]; copyValues( ds1, result ); return result; } case 2: { double[][] result = new double[ds1.length()][ds1.length(0)]; copyValues( ds1, result ); return result; } case 3: { double[][][] result = new double[ds1.length()][ds1.length(0)][ds1.length(0,0)]; copyValues( ds1, result ); return result; } default: throw new IllegalArgumentException("rank limit"); } } } /** * return an 1,2,or 3-D array of doubles or floats containing the values * in the default dataset. * @return */ public Object values() { return values(name); } /** * return an 1,2,or 3-D array of doubles or floats containing the values * in a slice on the zeroth dimension of the default dataset. * @param i0 the index to slice on. * @return 1,2,or 3-D array of doubles or floats. */ public Object slice(int i0 ) { return slice(name,i0); } /** * slice on the first dimension, which is useful for extracting * data component by component. * @param index * @return array of floats or doubles. */ public Object slice1( int index ) { return slice1( index, name ); } public Object slice1( int index, String name ) { throw new IllegalArgumentException("not implemented"); } public String depend(int dim) { QDataSet result = (QDataSet) this.ds.property("DEPEND_" + dim); if (result == null) return ""; else return nameFor(result); } public String plane(int iplane) { QDataSet result = (QDataSet) this.ds.property("PLANE_" + iplane); if (result == null) return ""; else return nameFor(result); } public String propertyAsString(String property) { Object result = this.ds.property(property); if (result == null) return ""; else return String.valueOf(result); } public double propertyAsDouble(String property) { Object result = this.ds.property(property); if (result == null) return Double.NaN; else return ((Number) result).doubleValue(); } public String propertyAsString(String name, String property) { Object result = datasets.get(name).property(property); if (result == null) return ""; else return String.valueOf(result); } public double propertyAsDouble(String name, String property) { Object result = datasets.get(name).property(property); if (result == null) return Double.NaN; else return ((Number) result).doubleValue(); } /** * copy the data into a 1-D array. Rank 2 and 3 dataset array are then * aliased with the highest dimension the most tightly packed. * @param name * @param result */ public void valuesAlias(String name, double[] result) { QDataSet ds1 = datasets.get(name); QubeDataSetIterator it = new QubeDataSetIterator(ds1); int iele = 0; while (it.hasNext()) { it.next(); result[iele] = it.getValue(ds1); iele++; } } /** * get the string values of the data instead of the numbers. This provides * a means to decode times and nominal data such as labels data sets. * @param name * @param result */ public void labelsAlias(String name, String[] result) { QDataSet ds1 = datasets.get(name); QubeDataSetIterator it = new QubeDataSetIterator(ds1); int iele = 0; Units u = (Units) ds1.property(QDataSet.UNITS); if (u == null) u = Units.dimensionless; while (it.hasNext()) { it.next(); result[iele] = u.createDatum(it.getValue(ds1)).toString(); iele++; } } /** * return the lengths of the dimensions of the specified dataset. * @param name * @return */ public int[] lengths(String name) { return DataSetUtil.qubeDims(datasets.get(name)); } /** * return the lengths of the dimensions of the default dataset. * @return */ public int[] lengths() { return DataSetUtil.qubeDims(datasets.get(name)); } /** * return the lengths of the dimensions of the named dataset, once sliced at index i. * @param name * @param i the index to slice * @return * @see #length() length which returns the length of the zeroth index. */ public int[] lengths(String name, int i) { QDataSet ds1= datasets.get(name); if ( ds1==null && sliceDep.containsKey(name) ) { ds1= (QDataSet) datasets.get(this.name).slice(i).property(sliceDep.get(name)); return DataSetUtil.qubeDims(ds1); } else { if ( ds1==null ) { throw new IllegalArgumentException("No such dataset: "+name); } return DataSetUtil.qubeDims(ds1.slice(i)); } } /** * return the lengths of the dimensions of the main dataset, once sliced at index i. * @param i the index to slice * @return the lengths of each dimension. */ public int[] lengths(int i) { return lengths(name,i); } /** * return the length of the zeroth dimension of the dataset * @param name * @return the length of the zeroth dimension */ public int length(String name) { QDataSet ds1= datasets.get(name); if ( ds1==null ) { throw new IllegalArgumentException("unable to get length for slice dataset, use lengths"); } else { return ds1.length(); } } /** * return the length of the zeroth dimension of the main dataset. * @return the length of the zeroth dimension */ public int length() { return length(name); } /** * return the number of dimensions of the specified dataset. * @param name * @return */ public int rank(String name) { return datasets.get(name).rank(); } /** * return the number of dimensions of the default dataset. * @return */ public int rank() { return datasets.get(name).rank(); } public boolean isQube() { return DataSetUtil.isQube( datasets.get(name) ); } /** * returns one of String, int, double, float, int[], double, float[] * @param name * @param propname * @param i * @return the property */ public Object property(String name, String propname, int i ) { Object prop = datasets.get(name).property(propname,i); if (prop instanceof QDataSet) { return nameFor((QDataSet) prop); } else if (prop instanceof Units) { for ( Units u: prefUnits ) { if ( u.isConvertibleTo( SemanticOps.getUnits( datasets.get(this.name) )) ) { return u.toString(); } } return prop.toString(); } else if ( propname.equals(QDataSet.FILL_VALUE) && this.useFill ) { return this.fill; } else { return prop; } } public boolean hasProperty(String name, String propname, int i) { return datasets.get(name).property(propname,i) != null; } /** * get the properties for the named dataset * @param name * @param i the index * @return */ public Map properties(String name,int i) { LinkedHashMap result = new LinkedHashMap<>(DataSetUtil.getProperties(datasets.get(name).slice(i))); //TODO: strange implementation for (String s : result.keySet()) { result.put(s, property(name, s,i)); } return result; } /** * returns one of String, int, double, float, int[], double, float[] * @param name * @param propname * @return */ public Object property(String name, String propname) { Object prop = datasets.get(name).property(propname); if (prop instanceof QDataSet) { return nameFor((QDataSet) prop); } else if (prop instanceof Units) { Units dsu= SemanticOps.getUnits( datasets.get(name) ); for ( Units u: prefUnits ) { if ( u.isConvertibleTo( dsu ) ) { return u.toString(); } } return prop.toString(); } else if ( propname.equals(QDataSet.FILL_VALUE) && this.useFill ) { return this.fill; } else { return prop; } } public boolean hasProperty(String name, String propname) { return datasets.get(name).property(propname) != null; } /** * get the properties for the named dataset * @param name * @return */ public Map properties(String name) { LinkedHashMap result = new LinkedHashMap<>(DataSetUtil.getProperties(datasets.get(name))); for (String s : result.keySet()) { result.put(s, property(name, s)); } return result; } /** * returns one of String, int, double, float, int[], double, float[] * @param propname * @return */ public Object property(String propname) { Object prop = datasets.get(name).property(propname); if (prop instanceof QDataSet) { return nameFor((QDataSet) prop); } else if (prop instanceof Units) { return prop.toString(); } else { return prop; } } public boolean hasProperty(String propname) { return datasets.get(name).property(propname) != null; } public Map properties() { LinkedHashMap result = new LinkedHashMap<>(DataSetUtil.getProperties(datasets.get(name))); for (String s : result.keySet()) { result.put(s, property(name, s)); } return result; } /** * returns one of String, int, double, float, int[], double, float[] * @param propname the property name * @param i the slice index * @return the name of the qdataset, or the value. */ public Object property(String propname, int i) { Object prop = datasets.get(name).property(propname,i); if (prop instanceof QDataSet) { return nameFor((QDataSet) prop); } else if (prop instanceof Units) { return prop.toString(); } else { return prop; } } public boolean hasProperty(String propname,int i) { return datasets.get(name).property(propname,i) != null; } public Map properties(int i) { LinkedHashMap result = new LinkedHashMap(DataSetUtil.getProperties(datasets.get(name).slice(i) )); return result; } public String[] names() { return datasets.keySet().toArray(new String[datasets.size()]); } public String name() { return name; } /** * print the Java memory stats to stderr. * @see #clearMemory() */ public void reportMemory() { System.err.println( "= Java Runtime Information =" ); String javaVersion= System.getProperty("java.version"); String javaVersionWarning= ""; // The java about checks for 1.8.102 String arch = System.getProperty("os.arch");// applet okay NumberFormat nf= new java.text.DecimalFormat(); String mem = nf.format( (Runtime.getRuntime()).maxMemory() / 1000000 ); String tmem= nf.format( (Runtime.getRuntime()).totalMemory() / 1000000 ); String fmem= nf.format( (Runtime.getRuntime()).freeMemory() / 1000000 ); System.err.println( "Java version: " + javaVersion + " " + javaVersionWarning ); System.err.println( "Arch: " + arch ); System.err.println( "Max memory (MB): " + mem + " (memory available to process)" ); System.err.println( "total memory (MB): " + tmem + " (amount allocated to the process)" ); System.err.println( "free memory (MB): " + fmem + " (amount available before more must be allocated)" ); } /** * return the total memory and free memory available to the Java process, * in megabytes (1e6 bytes). * @return an integer array [ used, total ] */ public int[] freeMemory() { int mem = (int)( (Runtime.getRuntime()).maxMemory() / 1000000 ); int tmem= (int)( (Runtime.getRuntime()).totalMemory() / 1000000 ); int fmem= (int)( (Runtime.getRuntime()).freeMemory() / 1000000 ); return new int[] { tmem-fmem, mem }; } /** * clear existing data from memory, in case the bridge object is not cleared * from in IDL or Matlab memory. * @see #reportMemory() */ public void clearMemory() { datasets.clear(); } }