Table of Contents
This algorithm converts adds/removes the normalisation to the fit function for calculating the TF asymmetry.
Name | Direction | Type | Default | Description |
---|---|---|---|---|
InputFunction | Input | Function | Mandatory | The fitting function to be converted. |
NormalizationTable | Input | TableWorkspace | Name of the table containing the normalizations for the asymmetries. | |
WorkspaceList | Input | str list | Mandatory | An ordered list of workspaces (to get the initial values for the normalizations). |
Mode | Input | string | Construct | Mode to run in. Construct will convert theinput function into one suitable for calculating the TF Asymmetry. Extract will find the original user function from a function that is suitable for TF Asymmetry calculations. Allowed values: [‘Extract’, ‘Construct’] |
OutputFunction | Output | Function | The converted fitting function. |
This algorithm can be run in two modes. The first is construct, which takes a user fitting function and it converts it to a TF normalisation function
where is the normalisation constant, is fixed to zero by default and is fixed to the Muon lifetime. The initial value for the normalisation constant is from the normalisation table.
The second mode is extract, if the TF normalisation function is given it will return the user function.
This algorithm works for both single and multi domain functions.
Example - Converting a function: This example is for converting a function.
import mantid.simpleapi as mantid
#create a normalisation table
tab = CreateEmptyTableWorkspace()
tab.addColumn('double', 'norm')
tab.addColumn('str', 'name')
tab.addColumn('str', 'method')
tab.addRow([1.,"Run;;Group;;a;;Asym;;#1","Estimate"])
tab.addRow([2.,"Run;;Group;;b;;Asym;;#1","Estimate"])
tab.addRow([3.,"Run;;Group;;c;;Asym;;#1","Estimate"])
tab.addRow([4.,"Run;;Group;;d;;Asym;;#1","Estimate"])
#create original function and workspace
myFunc='name=LinearBackground,A0=3,A1= 4;name=LinearBackground,A0=0,A1=2;ties=(f0.A1=3, f0.A0=f1.A0)'
ws = CreateWorkspace(DataX=[1,2,3,4,5,6,7], DataY=[1,2,3,4,5,6],OutputWorkspace="Run; Group; a; Asym; #1")
TFFunc = ConvertFitFunctionForMuonTFAsymmetry(InputFunction=myFunc, NormalizationTable=tab,
WorkspaceList=["Run; Group; a; Asym; #1"], Mode="Construct")
# do a fit with new function
fit =mantid.AlgorithmManager.create("Fit")
fit.setProperty("Function",str(TFFunc))
fit.setProperty("InputWorkspace",ws)
fit.setProperty("Output",'fitWS')
fit.execute()
fittedFunc = fit.getPropertyValue("Function")
returnFunc = ConvertFitFunctionForMuonTFAsymmetry(InputFunction=str(fittedFunc),NormalizationTable=tab,
WorkspaceList=["Run; Group; a; Asym; #1"], Mode="Extract")
# 0 iteration fit to get param table -> wont change function values
fit_output = Fit(Function=str(returnFunc),InputWorkspace=ws,MaxIterations=0,Output="return_params")
paramTable = fit_output.OutputParameters # table containing the optimal fit parameters
if paramTable.column(1)[0] == paramTable.column(1)[2]:
print("Constant tie has been preserved")
else:
print("Constant tie has not been preserved")
if paramTable.column(1)[1] == 3.0:
print("Fix has been preserved")
else:
print("Fix has not been presreved")
Output:
Constant tie has been preserved
Fix has been preserved
Categories: AlgorithmIndex | Muon
C++ source: ConvertFitFunctionForMuonTFAsymmetry.cpp (last modified: 2019-06-12)
C++ header: ConvertFitFunctionForMuonTFAsymmetry.h (last modified: 2019-10-28)