gCubeSystem
/
ecological-engine
16
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
ecological-engine/src/main/java/org/gcube/dataanalysis/ecoengine/evaluation/bioclimate/BioClimateAnalysis.java

759 lines
32 KiB
Java
Raw Blame History

package org.gcube.dataanalysis.ecoengine.evaluation.bioclimate;
import java.awt.Image;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.UUID;
import org.gcube.contentmanagement.graphtools.utils.MathFunctions;
import org.gcube.contentmanagement.lexicalmatcher.analysis.core.LexicalEngineConfiguration;
import org.gcube.dataanalysis.ecoengine.configuration.AlgorithmConfiguration;
import org.gcube.dataanalysis.ecoengine.datatypes.PrimitiveType;
import org.gcube.dataanalysis.ecoengine.interfaces.ComputationalAgent;
import org.gcube.dataanalysis.ecoengine.processing.factories.EvaluatorsFactory;
import org.gcube.dataanalysis.ecoengine.utils.DatabaseFactory;
import org.gcube.dataanalysis.ecoengine.utils.DatabaseUtils;
import org.gcube.dataanalysis.ecoengine.utils.Operations;
import org.hibernate.SessionFactory;
import org.jfree.data.category.DefaultCategoryDataset;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
public class BioClimateAnalysis {
private static Logger logger = LoggerFactory.getLogger(BioClimateAnalysis.class);
private static String countHighProbabilityCells = "select count(*) from %1$s where probability>%2$s";
private static String countSeaCells = "select count(*) from %1$s as a join %2$s as b on a.oceanarea>0 and a.csquarecode=b.csquarecode and ((a.iceconann<b.iceconann-%3$s or a.iceconann>b.iceconann+%3$s) or " + "(a.salinitymean<b.salinitymean-%3$s or a.salinitymean>b.salinitymean+%3$s) or (a.sstanmean<b.sstanmean-%3$s or a.sstanmean>b.sstanmean+%3$s))";
private static String iceLeakage = "select count(*) from %1$s as a join %2$s as b on a.csquarecode=b.csquarecode and (a.iceconann<b.iceconann) and a.oceanarea>0";
// private static String takeRangeOfDepths = "select distinct depthmin, max(depthmax) from %1$s group by depthmin order by depthmin";
private static String takeRangeOfDepths = "select distinct depthmin, depthmax from %1$s order by depthmin";
private static String takeRangeOfParam = "select %1$s from %2$s where %1$s is not null %3$s order by %1$s";
private static String countNumberOfSpeciesPerRange = "select count(*) from %1$s where %4$s>%2$s and %4$s<=%3$s ";
private static String countNumberOfSpecies = "select count(*) from %1$s where depthmin<%2$s and depthmin>=%3$s and depthmax<%4$s and depthmax>=%5$s";
private static String countProbabilityPerArea = "select count(*) from %1$s as a join hcaf_s as b on b.%2$s = %3$s and a.probability > #THRESHOLD# and a.csquarecode=b.csquarecode";
public static String salinityDefaultRange = "salinitymin>27.44 and salinitymin<=36.57";
public static String salinityMinFeature = "salinitymin";
private static enum FIELD {
iceconann, sstanmean, salinitymean
};
private static String takeAvgSelection = "select avg(%1$s),%2$s from %3$s %4$s group by %2$s order by %2$s";
private static String[] selectionCriteria = { "faoaream", "lme" };
private static String[] timeseriesNames = { "Ice Conc.", "Sea Surface Temperature", "Salinity" };
private static String[] quantitiesNames = { FIELD.iceconann.name(), FIELD.sstanmean.name(), FIELD.salinitymean.name() };
private static String[] criteriaFilters = { "where faoaream>0", "where lme>0" };
private static String[] criteriaNames = { "FaoArea", "LME" };
// private static String takeSubHspec = "select %1$s from #CLAUSE# order by %1$s";
private static String meanVal = "select avg(%2$s) from %1$s where oceanarea>0";
// private static String meanVal = "select %2$s from %1$s where csquarecode = '1311:478:4'";
private String configPath;
private String temporaryDirectory;
protected SessionFactory referencedbConnection;
private String[] csquareTable;
private String[] finalDistributionTable;
// hspec
private int[] highProbabilityCells;
private double[] discrepancies;
// hcaf
private double[] avgIce;
private double[] avgSST;
private double[] avgSalinity;
private ComputationalAgent eval;
private float status;
private boolean liveRender;
boolean doHcafAn;
boolean doHspecAn;
private LexicalEngineConfiguration config;
static int width = 680;
static int height = 420;
static int defaultNumOfFeatureClusters = 10;
public static void main(String[] args) throws Exception {
String configPath = "./cfg/";
String databaseUrl = "jdbc:postgresql://localhost/testdb";
String databaseUser = "gcube";
String databasePassword = "d4science2";
BioClimateAnalysis bioClimate = new BioClimateAnalysis(configPath, configPath, databaseUrl, databaseUser, databasePassword, true);
bioClimate.produceGraphs2D();
}
private static String[] SERIES = { "High Probability Cells Trend (>%1$s)", "Number of Changing Cells", "Reducing Ice Concentration Trend", "High Probability Cells Trend Derivative", "Average Trends", "Ice Concentration", "Sea Surface Temperature", "Salinity" };
public void produceGraphs2D() throws Exception {
DefaultCategoryDataset testpoints = new DefaultCategoryDataset();
// double[] points = Operations.parabolicInterpolation(-200d, 200d, 100);
double[] points = Operations.parabolicInterpolation(0.1926, 0.1727, 20);
// double[] points = Operations.inverseParabolicInterpolation(-200d, 300d, 50);
// double[] points = Operations.inverseExponentialInterpolation(29d, 30d, 10);
for (int i = 0; i < points.length; i++) {
// System.out.println(points[i]);
testpoints.addValue(points[i], "Points", "" + i);
}
BioClimateGraph lineg9 = new BioClimateGraph("parabols", Operations.getMax(points), Operations.getMin(points));
lineg9.render(testpoints);
}
private HashMap<String, Image> producedImages;
public HashMap<String, Image> getProducedImages() {
return producedImages;
}
private void produceGraphs(String[] csquareTableNames, String[] hspecTableNames, float threshold) throws Exception {
csquareTableNames = checkTableNames(csquareTableNames);
hspecTableNames = checkTableNames(hspecTableNames);
producedImages = new HashMap<String, Image>();
int numberOfTrends = highProbabilityCells.length;
// create the datasets...
DefaultCategoryDataset probabilityTrend = new DefaultCategoryDataset();
DefaultCategoryDataset discrepanciesTrend = new DefaultCategoryDataset();
DefaultCategoryDataset avgIceD = new DefaultCategoryDataset();
DefaultCategoryDataset avgSSTD = new DefaultCategoryDataset();
DefaultCategoryDataset avgSalinityD = new DefaultCategoryDataset();
for (int i = 0; i < numberOfTrends; i++) {
if (doHcafAn) {
avgIceD.addValue(avgIce[i], "Ice Conc.", csquareTableNames[i]);
avgSSTD.addValue(avgSST[i], "SST", csquareTableNames[i]);
avgSalinityD.addValue(avgSalinity[i], "Salinity", csquareTableNames[i]);
}
if (doHspecAn) {
probabilityTrend.addValue(highProbabilityCells[i], "Number Of Cells", hspecTableNames[i]);
if (i > 0) {
discrepanciesTrend.addValue(discrepancies[i], "Mean Discrepancy Respect to Prev. Distrib.", finalDistributionTable[i]);
}
}
}
if (doHspecAn) {
double min = Operations.getMin(discrepancies);
discrepancies[0] = min;
if (liveRender) {
BioClimateGraph lineg1 = new BioClimateGraph(String.format(SERIES[0], threshold), Operations.getMax(highProbabilityCells), Operations.getMin(highProbabilityCells));
BioClimateGraph lineg4 = new BioClimateGraph(SERIES[3], Operations.getMax(discrepancies), min);
lineg4.render(discrepanciesTrend);
lineg1.render(probabilityTrend);
}
producedImages.put("Probability_Trend", BioClimateGraph.renderStaticImgObject(width, height, probabilityTrend, String.format(SERIES[0], threshold), Operations.getMax(highProbabilityCells), Operations.getMin(highProbabilityCells)));
producedImages.put("Probability_Discrepancies_Trend", BioClimateGraph.renderStaticImgObject(width, height, discrepanciesTrend, SERIES[3], Operations.getMax(discrepancies), min));
}
if (doHcafAn) {
if (liveRender) {
BioClimateGraph lineg6 = new BioClimateGraph(SERIES[5], Operations.getMax(avgIce), Operations.getMin(avgIce));
BioClimateGraph lineg7 = new BioClimateGraph(SERIES[6], Operations.getMax(avgSST), Operations.getMin(avgSST));
BioClimateGraph lineg8 = new BioClimateGraph(SERIES[7], Operations.getMax(avgSalinity), Operations.getMin(avgSalinity));
lineg6.render(avgIceD);
lineg7.render(avgSSTD);
lineg8.render(avgSalinityD);
}
producedImages.put("Average_Ice_Concentration", BioClimateGraph.renderStaticImgObject(width, height, avgIceD, SERIES[5], Operations.getMax(avgIce), Operations.getMin(avgIce)));
producedImages.put("Average_SST", BioClimateGraph.renderStaticImgObject(width, height, avgSSTD, SERIES[6], Operations.getMax(avgSST), Operations.getMin(avgSST)));
producedImages.put("Average_Salinity", BioClimateGraph.renderStaticImgObject(width, height, avgSalinityD, SERIES[7], Operations.getMax(avgSalinity), Operations.getMin(avgSalinity)));
}
logger.trace("Produced All Images");
}
public void hcafEvolutionAnalysis(String[] hcafTable, String[] hcafTableNames) throws Exception {
globalEvolutionAnalysis(hcafTable, null, hcafTableNames, null, null, null, 0f);
}
public void hspecEvolutionAnalysis(String[] hspecTables, String[] hspecTableNames, String probabilityColumn, String csquareColumn, float threshold) throws Exception {
globalEvolutionAnalysis(null, hspecTables, null, hspecTableNames, probabilityColumn, csquareColumn, threshold);
}
private String[] checkTableNames(String[] tablesNames) {
ArrayList<String> newtables = new ArrayList<String>();
if ((tablesNames == null) || (tablesNames.length == 0))
return tablesNames;
for (String table : tablesNames) {
int i = 1;
String originalTable = table;
while (newtables.contains(table)) {
table = originalTable + "_" + i;
i++;
}
newtables.add(table);
}
String[] tables = new String[tablesNames.length];
for (int j = 0; j < tablesNames.length; j++) {
tables[j] = newtables.get(j);
}
return tables;
}
public void produceCharts(HashMap<String, HashMap<String, double[]>> GeoMap, String[] tablesNames) {
// produce a char for each feature
tablesNames = checkTableNames(tablesNames);
producedImages = new HashMap<String, Image>();
for (String featurename : GeoMap.keySet()) {
DefaultCategoryDataset chart = new DefaultCategoryDataset();
HashMap<String, double[]> timeseries = GeoMap.get(featurename);
double absmax = -Double.MAX_VALUE;
double absmin = Double.MAX_VALUE;
for (String timeserie : timeseries.keySet()) {
double[] points = timeseries.get(timeserie);
for (int i = 0; i < points.length; i++) {
if (points[i] > absmax)
absmax = points[i];
if (points[i] < absmin)
absmin = points[i];
chart.addValue(points[i], timeserie, tablesNames[i]);
}
}
if (liveRender) {
BioClimateGraph lineg1 = new BioClimateGraph(featurename, absmax, absmin);
lineg1.render(chart);
}
producedImages.put(featurename.replace(" ", "_"), BioClimateGraph.renderStaticImgObject(width, height, chart, featurename, absmax, absmin));
}
}
/**
* Generates a chart for hspens in time according to a certain interval in the parameter e.g. : a chart for several salinity intervals
*
* @param hspenTables
* @param hspenTableNames
* @param parameterName
* @param condition
* @throws Exception
*/
public void speciesEvolutionAnalysis(String[] hspenTables, String[] hspenTableNames, String parameterName, String condition) throws Exception {
try {
referencedbConnection = DatabaseFactory.initDBConnection(configPath + AlgorithmConfiguration.defaultConnectionFile, config);
logger.debug("ReferenceDB initialized");
status = 0f;
int numbOfTables = (hspenTables != null) ? hspenTables.length : 0;
if (numbOfTables > 0) {
// a map for each range of features: depth[0,100] ,depth [100,200]
HashMap<String, HashMap<String, double[]>> GeoMap = new HashMap<String, HashMap<String, double[]>>();
float statusstep = 80f / (float) numbOfTables;
if (condition != null && (condition.length() > 0))
condition = "and " + condition;
// take the spectrum of depths
logger.trace("Range query: " + String.format(takeRangeOfParam, parameterName, hspenTables[0], condition));
List<Object> paramrange = DatabaseFactory.executeSQLQuery(String.format(takeRangeOfParam, parameterName, hspenTables[0], condition), referencedbConnection);
int numberOfParams = paramrange.size();
double absolutePMin = Double.parseDouble("" + ((Object) paramrange.get(0)));
double absolutePMax = Double.parseDouble("" + ((Object) paramrange.get(numberOfParams - 1)));
double step = (absolutePMax - absolutePMin) / (double) defaultNumOfFeatureClusters;
int pClusters[] = new int[defaultNumOfFeatureClusters + 1];
pClusters[0] = 0;
for (int i = 1; i < pClusters.length; i++) {
double pToFind = absolutePMin + step * (i + 1);
int k = 0;
for (Object row : paramrange) {
if (Double.parseDouble("" + ((Object) row)) > pToFind)
break;
k++;
}
if (k >= numberOfParams)
k = numberOfParams - 1;
pClusters[i] = k;
}
// for each table
for (int i = 0; i < numbOfTables; i++) {
double pmax = 0;
// for each cluster build up a chart
for (int j = 1; j < pClusters.length; j++) {
double prevpmax = MathFunctions.roundDecimal(Double.parseDouble("" + (Object) paramrange.get(pClusters[j - 1])), 2);
pmax = MathFunctions.roundDecimal(Double.parseDouble("" + (Object) paramrange.get(pClusters[j])), 2);
if (prevpmax != pmax) {
// take the number of elements for this range
String countSpeciesQuery = String.format(countNumberOfSpeciesPerRange, hspenTables[i], prevpmax, pmax, parameterName);
logger.trace("count elements query: " + countSpeciesQuery);
List<Object> elementsInRange = DatabaseFactory.executeSQLQuery(countSpeciesQuery, referencedbConnection);
int nelements = (elementsInRange == null) ? 0 : Integer.parseInt("" + elementsInRange.get(0));
logger.trace("Number of elements for " + hspenTables[i] + " in (" + prevpmax + " - " + pmax + ")" + " : " + nelements);
// take the chart for this range
String chartName = parameterName + " envelope for interval (" + prevpmax + " ; " + pmax + ")";
// build the chart
HashMap<String, double[]> submap = GeoMap.get(chartName);
if (submap == null) {
submap = new HashMap<String, double[]>();
GeoMap.put(chartName, submap);
}
String timeseries = "number of species";
double[] elements = submap.get(timeseries);
if (elements == null) {
elements = new double[numbOfTables];
submap.put(timeseries, elements);
}
elements[i] = nelements;
}
}
status = status + statusstep;
}
status = 80f;
produceCharts(GeoMap, hspenTableNames);
}
} catch (Exception e) {
e.printStackTrace();
throw e;
} finally {
status = 100f;
try{
referencedbConnection.close();
}catch(Exception e){}
}
}
/**
* Generates a chart for hspec probability > thr in each Fao Area and LME
*
* @param hspecTables
* @param hspecTablesNames
* @throws Exception
*/
public void speciesGeographicEvolutionAnalysis2(String[] hspecTables, String[] hspecTablesNames, float threshold) throws Exception {
try {
referencedbConnection = DatabaseFactory.initDBConnection(configPath + AlgorithmConfiguration.defaultConnectionFile, config);
logger.debug("ReferenceDB initialized");
status = 0f;
int numbOfTables = (hspecTables != null) ? hspecTables.length : 0;
if (numbOfTables > 0) {
// a map for each feature. each sub map contains a trend for faoaream, lme etc.
HashMap<String, HashMap<String, double[]>> GeoMap = new HashMap<String, HashMap<String, double[]>>();
float statusstep = 80f / (float) numbOfTables;
// for each table
for (int i = 0; i < numbOfTables; i++) {
// for each criterion to apply: fao area, lme etc.
for (int j = 0; j < criteriaNames.length; j++) {
List<Object> listCriterion = DatabaseFactory.executeSQLQuery(DatabaseUtils.getDinstictElements("hcaf_s", selectionCriteria[j], criteriaFilters[j]), referencedbConnection);
for (Object code : listCriterion) {
String code$ = "" + code;
String query = String.format(countProbabilityPerArea, hspecTables[i], selectionCriteria[j], code$);
query = query.replace("#THRESHOLD#", "" + threshold);
logger.trace("Executing query for counting probabilities: " + query);
List<Object> counts = DatabaseFactory.executeSQLQuery(query, referencedbConnection);
logger.trace("Query Executed");
int countPerArea = (counts == null) ? 0 : Integer.parseInt("" + counts.get(0));
String chartName = "Hspec (prob>0.8) for " + criteriaNames[j] + "_" + code$; // put the code and the value in the timeseries associated to the feature name
HashMap<String, double[]> submap = GeoMap.get(chartName);
if (submap == null) {
submap = new HashMap<String, double[]>();
GeoMap.put(chartName, submap);
}
String timeseries = "number of occupied cells";
double[] elements = submap.get(timeseries);
if (elements == null) {
elements = new double[numbOfTables];
submap.put(timeseries, elements);
}
elements[i] = countPerArea;
}
}
status = status + statusstep;
}
status = 80f;
produceCharts(GeoMap, hspecTablesNames);
}
} catch (Exception e) {
e.printStackTrace();
throw e;
} finally {
status = 100f;
referencedbConnection.close();
}
}
public void speciesGeographicEvolutionAnalysis(String[] hspecTables, String[] hspecTablesNames, float threshold) throws Exception {
try {
referencedbConnection = DatabaseFactory.initDBConnection(configPath + AlgorithmConfiguration.defaultConnectionFile, config);
logger.debug("ReferenceDB initialized");
status = 0f;
int numbOfTables = (hspecTables != null) ? hspecTables.length : 0;
if (numbOfTables > 0) {
// a map for each feature. each sub map contains a trend for faoaream, lme etc.
HashMap<String, HashMap<String, double[]>> GeoMap = new HashMap<String, HashMap<String, double[]>>();
float statusstep = 80f / (float) numbOfTables;
// for each table
for (int i = 0; i < numbOfTables; i++) {
String tmpanalysisTable = "tmpanalysis" + ("" + UUID.randomUUID()).replace("-", "").replace("_", "");
try {
DatabaseFactory.executeSQLUpdate("drop table " + tmpanalysisTable, referencedbConnection);
} catch (Exception ee) {
logger.trace("table " + tmpanalysisTable + " does not exist");
}
String preparationQuery = "create table " + tmpanalysisTable + " as select a.faoaream, lme,count(*) from %1$s as a where a.probability > #THRESHOLD# group by faoaream,lme;";
preparationQuery = String.format(preparationQuery, hspecTables[i]);
preparationQuery = preparationQuery.replace("#THRESHOLD#", "" + threshold);
logger.trace("Executing query for counting probabilities: " + preparationQuery);
DatabaseFactory.executeSQLUpdate(preparationQuery, referencedbConnection);
logger.trace("Query Executed");
// for each criterion to apply: fao area, lme etc.
for (int j = 0; j < criteriaNames.length; j++) {
String criteriaQuery = String.format("select %1$s,sum(count) from " + tmpanalysisTable + " %2$s group by %1$s;", selectionCriteria[j], criteriaFilters[j]);
logger.trace("Executing query for counting probabilities: " + criteriaQuery);
List<Object> codeSums = DatabaseFactory.executeSQLQuery(criteriaQuery, referencedbConnection);
for (Object codeSum : codeSums) {
String code$ = "" + ((Object[]) codeSum)[0];
int countPerArea = (((Object[]) codeSum)[1] == null) ? 0 : Integer.parseInt("" + ((Object[]) codeSum)[1]);
logger.trace("Analyzing " + selectionCriteria[j] + " with code " + code$ + " count " + countPerArea);
String chartName = "Hspec (prob>0.8) for " + criteriaNames[j] + "_" + code$;
// put the code and the value in the timeseries associated to the feature name
HashMap<String, double[]> submap = GeoMap.get(chartName);
if (submap == null) {
submap = new HashMap<String, double[]>();
GeoMap.put(chartName, submap);
}
String timeseries = "number of occupied cells";
double[] elements = submap.get(timeseries);
if (elements == null) {
elements = new double[numbOfTables];
submap.put(timeseries, elements);
}
elements[i] = countPerArea;
}
}
try {
DatabaseFactory.executeSQLUpdate("drop table " + tmpanalysisTable, referencedbConnection);
} catch (Exception ee) {
ee.printStackTrace();
logger.trace("table " + tmpanalysisTable + " does not exist");
}
status = status + statusstep;
}
status = 80f;
produceCharts(GeoMap, hspecTablesNames);
}
} catch (Exception e) {
e.printStackTrace();
throw e;
} finally {
status = 100f;
referencedbConnection.close();
}
}
/**
* Generates a geographic trend for each hspec feature: ice con, salinity, sst in each fao area
*
* @param hcafTable
* @param hcafTableNames
* @throws Exception
*/
public void geographicEvolutionAnalysis(String[] hcafTable, String[] hcafTableNames) throws Exception {
try {
referencedbConnection = DatabaseFactory.initDBConnection(configPath + AlgorithmConfiguration.defaultConnectionFile, config);
logger.debug("ReferenceDB initialized");
doHcafAn = true;
doHspecAn = true;
status = 0f;
this.csquareTable = hcafTable;
int numbOfTables = (hcafTable != null) ? hcafTable.length : 0;
if (numbOfTables > 0) {
// a map for each feature. each sub map contains a trend for faoaream, lme etc.
HashMap<String, HashMap<String, double[]>> GeoMap = new HashMap<String, HashMap<String, double[]>>();
float statusstep = 80f / (float) numbOfTables;
// for each table
for (int i = 0; i < numbOfTables; i++) {
// for each criterion to apply: fao area, lme etc.
for (int j = 0; j < criteriaNames.length; j++) {
// for each quantity to display: ice concentration
for (int k = 0; k < quantitiesNames.length; k++) {
String query = String.format(takeAvgSelection, quantitiesNames[k], selectionCriteria[j], hcafTable[i], criteriaFilters[j]);
logger.debug("Query to be executed : " + query);
// take couples (avg,code)
List<Object> quantityCriterion = DatabaseFactory.executeSQLQuery(query, referencedbConnection);
// for each row
for (Object element : quantityCriterion) {
Object[] row = (Object[]) element;
// take avg value
double value = (row[0] == null) ? 0 : Double.parseDouble("" + row[0]);
// take code for criterion
String code = "" + row[1];
String chartName = timeseriesNames[k] + " for " + criteriaNames[j] + "_" + code;
// put the code and the value in the timeseries associated to the feature name
HashMap<String, double[]> submap = GeoMap.get(chartName);
if (submap == null) {
submap = new HashMap<String, double[]>();
GeoMap.put(chartName, submap);
}
String timeseries = criteriaNames[j] + "_" + code;
double[] elements = submap.get(timeseries);
if (elements == null) {
elements = new double[numbOfTables];
submap.put(timeseries, elements);
}
elements[i] = value;
}
}
}
status = status + statusstep;
}
status = 80f;
produceCharts(GeoMap, hcafTableNames);
}
} catch (Exception e) {
e.printStackTrace();
throw e;
} finally {
status = 100f;
referencedbConnection.close();
}
}
/**
* Generates a chart for each hspec feature Generates a chart for hspec prob > thr and performs a discrepancy analysis on hspec
*
* @param hcafTable
* @param hspecTables
* @param hcafTablesNames
* @param hspecTableNames
* @param probabilityColumn
* @param csquareColumn
* @throws Exception
*/
public void globalEvolutionAnalysis(String[] hcafTable, String[] hspecTables, String[] hcafTablesNames, String[] hspecTableNames, String probabilityColumn, String csquareColumn, float threshold) throws Exception {
try {
referencedbConnection = DatabaseFactory.initDBConnection(configPath + AlgorithmConfiguration.defaultConnectionFile, config);
logger.debug("ReferenceDB initialized");
doHcafAn = true;
doHspecAn = true;
if (hcafTable == null)
doHcafAn = false;
if (hspecTables == null)
doHspecAn = false;
status = 0f;
this.csquareTable = hcafTable;
this.finalDistributionTable = hspecTables;
int numbOfPoints = (hcafTable != null) ? hcafTable.length : hspecTables.length;
highProbabilityCells = new int[numbOfPoints];
discrepancies = new double[numbOfPoints];
avgIce = new double[numbOfPoints];
avgSST = new double[numbOfPoints];
avgSalinity = new double[numbOfPoints];
float statusstep = 80f / (float) numbOfPoints;
// create temp table puppa as select count(*), probability > 0.8 as aboveThreshold, sum(probability) as partialprobability from hspec_2050_suitable group by probability >0.8;
// select count(*), probability > 0.8 as aboveThreshold, sum(probability) as partialprobability from hspec_2050_suitable group by probability >0.8;
// select sum(count) as count, 3 as x from puppa union select count,2 as x from puppa where abovethreshold = true union select sum(partialprobability) as count, 1 as x from puppa order by x desc;
for (int i = 0; i < numbOfPoints; i++) {
if (doHspecAn)
highProbabilityCells[i] = calcHighProbabilityCells(hspecTables[i], threshold);
if (doHcafAn) {
avgIce[i] = avgValue(hcafTable[i], FIELD.iceconann.name());
avgSST[i] = avgValue(hcafTable[i], FIELD.sstanmean.name());
avgSalinity[i] = avgValue(hcafTable[i], FIELD.salinitymean.name());
logger.trace("(" + hcafTable[i] + "): " + " ICE " + avgIce[i] + " SST " + avgSST[i] + " SAL " + avgSalinity[i]);
}
if (doHspecAn) {
if (i == 0) {
discrepancies[i] = 1.0;
} else {
// OLD CALCULATION discrepancies[i] = MathFunctions.roundDecimal(calcDiscrepancy(configPath, temporaryDirectory, hspecTables[i], hspecTables[i - 1], probabilityColumn, csquareColumn, 0.1f), 5);
// discrepancies[i] = MathFunctions.roundDecimal(calcOverDiscrepancy(configPath, temporaryDirectory, hspecTables[i], hspecTables[i - 1], probabilityColumn, csquareColumn, 0.1f), 5);
discrepancies[i] = highProbabilityCells[i]-highProbabilityCells[i-1];
}
logger.trace("(" + hspecTables[i] + "): DISCREPANCY " + discrepancies[i] + " HIGH PROB CELLS " + highProbabilityCells[i]);
}
// logger.trace("(" + hcafTable[i] + "," + hspecTables[i] + "): HIGH PROB CELLS " + highProbabilityCells[i] + " DISCREPANCY " + discrepancies[i] + " ICE " + avgIce[i] + " SST " + avgSST[i] + " SAL " + avgSalinity[i]);
status = status + statusstep;
}
status = 80f;
produceGraphs(hcafTablesNames, hspecTableNames, threshold);
} catch (Exception e) {
e.printStackTrace();
throw e;
} finally {
status = 100f;
try{
referencedbConnection.close();
}catch(Exception e2){}
}
}
// init connections
public BioClimateAnalysis(String configPath, String persistencePath, String databaseURL, String databaseUserName, String databasePassword, boolean liveRender) throws Exception {
this.configPath = configPath;
this.temporaryDirectory = persistencePath;
if (!configPath.endsWith("/"))
configPath += "/";
if (!persistencePath.endsWith("/"))
this.temporaryDirectory += "/";
config = new LexicalEngineConfiguration();
config.setDatabaseURL(databaseURL);
config.setDatabaseUserName(databaseUserName);
config.setDatabasePassword(databasePassword);
this.liveRender = liveRender;
}
public int calcHighProbabilityCells(String hspec, double probabilty) throws Exception {
logger.trace("Calculating High Prob Cells: "+String.format(countHighProbabilityCells, hspec, probabilty));
List<Object> countage = DatabaseFactory.executeSQLQuery(String.format(countHighProbabilityCells, hspec, probabilty), referencedbConnection);
int count = Integer.parseInt("" + countage.get(0));
logger.trace("Calc High Prob Cells: " + count);
return count;
}
public double avgValue(String hcaf1, String field) throws Exception {
List<Object> countage = DatabaseFactory.executeSQLQuery(String.format(meanVal, hcaf1, field), referencedbConnection);
double count = 0;
if (countage != null && countage.size() > 0)
count = Double.parseDouble("" + countage.get(0));
return count;
}
public int countIceLeakageCells(String hcaf1, String hcaf2) throws Exception {
List<Object> countage = DatabaseFactory.executeSQLQuery(String.format(iceLeakage, hcaf1, hcaf2), referencedbConnection);
int count = Integer.parseInt("" + countage.get(0));
return count;
}
public int countSeaCells(String hcaf1, String hcaf2, double threshold) throws Exception {
// System.out.println(String.format(countSeaCells, hcaf1, hcaf2,threshold));
List<Object> countage = DatabaseFactory.executeSQLQuery(String.format(countSeaCells, hcaf1, hcaf2, threshold), referencedbConnection);
int count = Integer.parseInt("" + countage.get(0));
return count;
}
public float getStatus() {
return status;
}
public double calcOverDiscrepancy(String configPath, String persistencePath, String firstTable, String secondTable, String probabilityColumnName, String csquareColumn, float comparisonThreshold) throws Exception {
List<Object> nelementsQ1 = DatabaseFactory.executeSQLQuery(DatabaseUtils.countElementsStatement(firstTable), referencedbConnection);
int nelements = Integer.parseInt("" + nelementsQ1.get(0));
logger.trace("Number Of elements1: " + nelementsQ1);
List<Object> nelementsQ2 = DatabaseFactory.executeSQLQuery(DatabaseUtils.countElementsStatement(secondTable), referencedbConnection);
int nelements2 = Integer.parseInt("" + nelementsQ2.get(0));
logger.trace("Number Of elements2: " + nelementsQ1);
List<Object> sumFirst = DatabaseFactory.executeSQLQuery(DatabaseUtils.sumElementsStatement(firstTable, probabilityColumnName), referencedbConnection);
double sum1 = Double.parseDouble("" + sumFirst.get(0));
logger.trace("Sum1: " + sum1);
List<Object> sumSecond = DatabaseFactory.executeSQLQuery(DatabaseUtils.sumElementsStatement(secondTable, probabilityColumnName), referencedbConnection);
double sum2 = Double.parseDouble("" + sumSecond.get(0));
logger.trace("Sum2: " + sum1);
double d = (double) (sum2 - sum1) / (double) (nelements + nelements2);
return d;
}
public double calcDiscrepancy(String configPath, String persistencePath, String firstTable, String secondTable, String probabilityColumnName, String csquareColumn, float comparisonThreshold) throws Exception {
AlgorithmConfiguration config = new AlgorithmConfiguration();
config.setConfigPath(configPath);
config.setPersistencePath(persistencePath);
config.setNumberOfResources(1);
config.setAgent("DISCREPANCY_ANALYSIS");
config.setParam("FirstTable", firstTable);
config.setParam("SecondTable", secondTable);
config.setParam("FirstTableCsquareColumn", csquareColumn);
config.setParam("SecondTableCsquareColumn", csquareColumn);
config.setParam("FirstTableProbabilityColumn", probabilityColumnName);
config.setParam("SecondTableProbabilityColumn", probabilityColumnName);
config.setParam("ComparisonThreshold", "" + comparisonThreshold);
config.setParam("MaxSamples", "" + 30000);
eval = EvaluatorsFactory.getEvaluators(config).get(0);
eval.compute();
PrimitiveType output = (PrimitiveType) eval.getOutput();
HashMap<String, String> out = (HashMap<String, String>)output.getContent();
Double d = Double.parseDouble(out.get("MEAN"));
return d;
}
}
View Git Blame Copy Permalink