iceCube.uhe.analysis
Class NPEthresholdOptimizer

java.lang.Object
  iceCube.uhe.analysis.NPEthresholdOptimizer

public class NPEthresholdOptimizer
extends java.lang.Object

Optimize NPE threshold to minimize either the model rejection potential or the signal discovery potential. Rely on numRecipes.FeldmanCousins.java Written originally by S. Yoshida for the IceCube EHE analysis. 2010/3/29

Field Summary

protected java.lang.String	aidaFileName name of the output Aida Tree file
protected IHistogram1D	bgRate1D
protected IHistogram2D	bgRate2D histogram to store number of background events
protected int	dimensionNPE dimension of the logNPE.
protected int	dimensionY dimension of the parameter y.
protected double	logNPEMax maximum range of logNPE
protected double	logNPEMin mininum range of logNPE
protected IHistogram1D	modelRejectHisto Histogram to store the model rejection factor
protected IHistogram1D	signalDiscoverPotentialHisto Histogram to store the signal discovery potential
protected IHistogram1D	signalRate1D
protected IHistogram2D	signalRate2D histogram to store number of signal events
protected double	yMax maximum range of the parameter y (ex.
protected double	yMin mininum range of the parameter y (ex.

Constructor Summary

NPEthresholdOptimizer()
Constructor.

NPEthresholdOptimizer(int dimensionY, double yMin, double yMax, int dimensionNPE, double logNPEMin, double logNPEMax)
Constructor.

Method Summary

void	calculateAndFillModelRejectionFactor(double y, boolean isIntegral) Calculate the model rejection factor for each NPE threshold and fill the 1D histogram(ogNPE) with the factors.
void	calculateAndFillSignalDiscoveryPotential(double y, double significanceOfDiscovery, boolean isIntegral) Calculate the signal discovery potential for each NPE threshold and fill the 1D histogram(ogNPE) with the factors.
static void	main(java.lang.String[] args) A simple main method for test
void	plot() Plot the model rejection factor and the signal discovery poteintial as a function of logNPE
void	readAndFillEventRate(java.io.DataInputStream in) Reading the number of signal and background events binned in logNPE and y(such as cos(Zenith)
void	readAndFillEventRate(java.io.DataInputStream in, boolean wideFormat) Reading the number of signal and background events binned in logNPE and y(such as cos(Zenith).

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

dimensionY

protected int dimensionY

dimension of the parameter y. usually cosZenith, but depends on analysis

yMax

protected double yMax

maximum range of the parameter y (ex. cos(Zenith))

yMin

protected double yMin

mininum range of the parameter y (ex. cos(Zenith))

dimensionNPE

protected int dimensionNPE

dimension of the logNPE. bin

logNPEMax

protected double logNPEMax

maximum range of logNPE

logNPEMin

protected double logNPEMin

mininum range of logNPE

signalRate2D

protected IHistogram2D signalRate2D

histogram to store number of signal events

signalRate1D

protected IHistogram1D signalRate1D

bgRate2D

protected IHistogram2D bgRate2D

histogram to store number of background events

bgRate1D

protected IHistogram1D bgRate1D

modelRejectHisto

protected IHistogram1D modelRejectHisto

Histogram to store the model rejection factor

signalDiscoverPotentialHisto

protected IHistogram1D signalDiscoverPotentialHisto

Histogram to store the signal discovery potential

aidaFileName

protected java.lang.String aidaFileName

name of the output Aida Tree file

Constructor Detail

NPEthresholdOptimizer

public NPEthresholdOptimizer(int dimensionY,
                             double yMin,
                             double yMax,
                             int dimensionNPE,
                             double logNPEMin,
                             double logNPEMax)
                      throws java.io.IOException

Constructor. Creaate the Jaida Histogram [logNPE][y] to store number of background and signal events.

        int dimensionY   dimension of the parameter y (ex. cos(zenith) in histograms
        double yMin      mininum rang of the parameter y (ex. cos(zenith) in histograms
        double yMax      maximum rang of the parameter y (ex. cos(zenith) in histograms
        int dimensionNPE dimension of log(NPE)  in histograms
        double logNPEMin    mininum rang of log(NPE) in histograms
        double logNPEMax    maximum rang of log(NPE) in histograms
        

Throws:

java.io.IOException

NPEthresholdOptimizer

public NPEthresholdOptimizer()
                      throws java.io.IOException

Constructor. Creaate the Jaida Histogram [logNPE][y] to store number of background and signal events. Range of bins and dimensions are set by default values

Throws:

java.io.IOException

Method Detail

readAndFillEventRate

public void readAndFillEventRate(java.io.DataInputStream in)
                          throws java.io.IOException

Reading the number of signal and background events binned in logNPE and y(such as cos(Zenith)

Throws:

java.io.IOException

readAndFillEventRate

public void readAndFillEventRate(java.io.DataInputStream in,
                                 boolean wideFormat)
                          throws java.io.IOException

Reading the number of signal and background events binned in logNPE and y(such as cos(Zenith). Data format is different from the method above.

Throws:

java.io.IOException

calculateAndFillModelRejectionFactor

public void calculateAndFillModelRejectionFactor(double y,
                                                 boolean isIntegral)

Calculate the model rejection factor for each NPE threshold and fill the 1D histogram(ogNPE) with the factors. You have to fill the background and signal number historam first by FillSignalRate() and FillBGRate()

        double y  : presumably cosZenith, a parameter on y-axis. logNPE is on x-axis.
        boolean isIntegral  : if the filled BG/Signal rate is integral rate, set true.
        

calculateAndFillSignalDiscoveryPotential

public void calculateAndFillSignalDiscoveryPotential(double y,
                                                     double significanceOfDiscovery,
                                                     boolean isIntegral)
                                              throws java.io.IOException

Calculate the signal discovery potential for each NPE threshold and fill the 1D histogram(ogNPE) with the factors. You have to fill the background and signal number historam first by FillSignalRate() and FillBGRate()

        double y  : presumably cosZenith, a parameter on y-axis. logNPE is on x-axis.
        double significanceOfDiscovery : significance to define "discovery" of signals. 5.0 e.x., 5sigma
        boolean isIntegral  : if the filled BG/Signal rate is integral rate, set true.
        

Throws:

java.io.IOException

plot

public void plot()

Plot the model rejection factor and the signal discovery poteintial as a function of logNPE

main

public static void main(java.lang.String[] args)
                 throws java.io.IOException

A simple main method for test

Throws:

java.io.IOException