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

EnggFitTOFFromPeaks v1

Summary

The algorithm fits an expected diffraction pattern to a workpace spectrum by performing single peak fits.

See Also

EnggFitPeaks, GSASIIRefineFitPeaks, Fit

Properties

Name

Direction

Type

Default

Description

FittedPeaks

Input

TableWorkspace

Mandatory

Information on fitted peaks, in the format produced by EnggFitPeaks. The table must contain, for every peak fitted the expected peak value (in d-spacing), and the parameters fitted. The expected values are given in the column labelled ‘dSpacing’. When using the back-to-back exponential peak function, the ‘X0’ column must have the fitted peak center.

OutParametersTable

Input

string

Name for a table workspace with the fitted values calculated by this algorithm (DIFA, DIFC and TZERO calibration parameters) for GSAS. These three parameters are added as three columns in a single row. If not given, the table workspace is not created.

DIFA

Output

number

Fitted DIFA calibration parameter

DIFC

Output

number

Fitted DIFC calibration parameter

TZERO

Output

number

Fitted TZERO calibration parameter

Description

Warning

This algorithm is being developed for a specific instrument. It might undergo significant changes, should instrument scientists decide to do so.

This algorithm finds the calibration parameters DIFA, DIFC and TZERO (as defined in GSAS) from a list of peaks fitted to a diffraction pattern. The peaks can be fitted to a Mantid workspace spectrum using the algorithm EnggFitPeaks which produces a table with parameter values for peaks in time-of-flight (TOF, see Unit Factory). The table is the essential input to this algorithm.

This algorithm fits the adjusted peak time-of-flight value positions that are fitted against expected dSpacing values (d) according to the expression:

\[TOF = DIFA*d^2 + DIFC*d + TZERO\]

The calibration parameters DIFA, TZERO and DIFC can then be used within the GSAS program or in other Mantid algorithms (see EnggCalibrate and EnggCalibrateFull). These parameters are returned and can be retrieved as output properties as well.

If a name is given in OutParametersTable this algorithm also produces a table workspace with that name, containing the parameters fitted (DIFA, DIFC, TZERO).

The parameters DIFA, DIFC, TZERO are also used in other Mantid algorithms. For example see AlignDetectors where these parameters are used to convert units from time-of-flight to d-spacing.

Usage

Example - Fitting DIFC parameter with two peaks:

# Three Back2Back exponential peaks
peak1 = "name=BackToBackExponential,I=6000,A=0.05,B=0.025,X0=15000,S=100"
peak2 = "name=BackToBackExponential,I=6000,A=0.05,B=0.025,X0=27500,S=100"
peak3 = "name=BackToBackExponential,I=5000,A=0.05,B=0.025,X0=35000,S=100"
bg = "name=FlatBackground,A0=20"

# Create workpsace with the above peaks and a single detector pixel
ws = CreateSampleWorkspace(Function="User Defined",
                          UserDefinedFunction=";".join([peak1, peak2, peak3, bg]),
                          NumBanks=1,
                          BankPixelWidth=1,
                          XMin=6000,
                          XMax=45000,
                          BinWidth=10)

# Update instrument geometry to something that would allow converting to some sane dSpacing values
EditInstrumentGeometry(Workspace = ws, L2 = [1.5], Polar = [90], PrimaryFlightPath = 50)

# Run the algorithm. Defaults are shown below. Files entered must be in .csv format and if both ExpectedPeaks and ExpectedPeaksFromFile are entered, the latter will be used.

peaks_tbl = EnggFitPeaks(ws, 0, [0.8, 1.5, 1.9])
out_tbl_name = 'difc_from_peaks'
difa, difc, tzero = EnggFitTOFFromPeaks(FittedPeaks=peaks_tbl, OutParametersTable=out_tbl_name)

# Print the results
print("DIFA: %.1f" % difa)
print("DIFC: %.0f" % round(difc,-1))
print("TZERO: %.0f" % round(tzero,-1))
tbl = mtd[out_tbl_name]
print("The output table has %d row(s)" % tbl.rowCount())
print("Number of peaks fitted: {0}".format(peaks_tbl.rowCount()))

Output:

DIFA: ...
DIFC: ...
TZERO: ...
The output table has 1 row(s)
Number of peaks fitted: 3

Categories: AlgorithmIndex | Diffraction\Engineering | Diffraction\Fitting

Source

Python: EnggFitTOFFromPeaks.py