\(\renewcommand\AA{\unicode{x212B}}\)

IntegratePeakTimeSlices v1

Warning

IntegratePeakTimeSlices is deprecated (on 2024-10-02) and has no replacement.

Summary

The algorithm uses CurveFitting::BivariateNormal for fitting a time slice

See Also

PeakIntegration

Properties

Name

Direction

Type

Default

Description

InputWorkspace

Input

MatrixWorkspace

Mandatory

A 2D workspace with X values of time of flight

OutputWorkspace

Output

TableWorkspace

Mandatory

Name of the output table workspace with Log info

Peaks

Input

PeaksWorkspace

Mandatory

Workspace of Peaks

PeakIndex

Input

number

0

Index of peak in PeaksWorkspace to integrate

PeakQspan

Input

number

0.06

Max magnitude of Q of Peak to Q of Peak Center, where mod(Q)=1/d

CalculateVariances

Input

boolean

True

Calc (co)variances given parameter values versus fit (co)Variances

Ties

Input

string

Tie parameters(Background,Intensity, Mrow,…) to values/formulas.

NBadEdgePixels

Input

number

0

Number of bad Edge Pixels

Intensity

Output

number

Peak Integrated Intensity

SigmaIntensity

Output

number

Peak Integrated Intensity Error

Description

This algorithm fits a bivariate normal distribution( plus background) to the data on each time slice. The Fit program uses BivariateNormal for the Fit Function.

The area used for the fitting is calculated based on the dQ parameter. A good value for dQ is 1/largest unit cell length. This parameter dictates the size of the area used to approximate the intensity of the peak. The estimate .1667/ max(a,b,c) assumes |Q|=1/d.

The result is returned in this algorithm’s output “Intensity” and “SigmaIntensity” properties. The peak object is NOT CHANGED.

The table workspace is also a result. Each line contains information on the fit for each good time slice. The column names( and information) in the table are:

`` Time, Channel, Background, Intensity, Mcol, Mrow, SScol,SSrow, SSrc, NCells,``
`` ChiSqrOverDOF, TotIntensity, BackgroundError, FitIntensityError, ISAWIntensity,``
``  ISAWIntensityError,TotalBoundary, NBoundaryCells, Start Row, End Row,Start Col, and End Col.``
``  The last column has a comma separated List of sepctral ID’s used in the time slice.``

The final Peak intensity is the sum of the IsawIntensity for each time slice. The error is the square root of the sum of squares of the IsawIntensityError values.

The columns whose names are Background, Intensity, Mcol, Mrow, SScol, SSrow, and SSrc correspond to the parameters for the BivariateNormal curve fitting function.

This algorithm has been carefully tweaked to give good results for interior peaks only. Peaks close to the edge of the detector may not give good results.

This Algorithm is also used by the PeakIntegration v1 algorithm when the Fit tag is selected.

Usage

# Load a SCD data set from systemtests Data and find the peaks
LoadEventNexus(Filename=r'TOPAZ_3132_event.nxs',OutputWorkspace='TOPAZ_3132_nxs')
ConvertToDiffractionMDWorkspace(InputWorkspace='TOPAZ_3132_nxs',OutputWorkspace='TOPAZ_3132_md',LorentzCorrection='1')
FindPeaksMD(InputWorkspace='TOPAZ_3132_md',PeakDistanceThreshold='0.15',MaxPeaks='100',OutputWorkspace='peaks')
FindUBUsingFFT(PeaksWorkspace='peaks',MinD='2',MaxD='16')
IndexPeaks(PeaksWorkspace='peaks')

# Run the Integration algorithm and print results
peak0,int,sig = IntegratePeakTimeSlices(InputWorkspace='TOPAZ_3132_nxs', Peaks='peaks', Intensity=4580.8587719746683, SigmaIntensity=190.21154129339735)
print("Intensity and SigmaIntensity of peak 0  = {} {}".format(int, sig))

Categories: AlgorithmIndex | Crystal\Integration | Deprecated

Source

C++ header: IntegratePeakTimeSlices.h

C++ source: IntegratePeakTimeSlices.cpp