The performance of the TS1 methane moderator is currently tracked by fitting peaks from the incident monitor on PEARL. This analysis is run manually within the Mantid software and involves maintaining lists of runs and cycle names to feed into the script, along with executing the actual code.
This procedure can be automated and the output data should be ingested into the data lakehouse to enabled tools such as Superset to view the corresponding plots.
The basics of the analysis amount to the following script:
Script
for i in range(start,end+1):
if i == 95382:
continue
Load(Filename='<file-path>', OutputWorkspace=str(i))
ws = mtd[str(i)]
run = ws.getRun()
pcharge = run.getProtonCharge()
if pcharge <1.0:
reject.append(str(i))
DeleteWorkspace(str(i))
continue
NormaliseByCurrent(InputWorkspace=str(i), OutputWorkspace=str(i))
ExtractSingleSpectrum(InputWorkspace=str(i),WorkspaceIndex=index,
OutputWorkspace=str(i)+ '_' + str(index))
CropWorkspace(InputWorkspace=str(i)+ '_' + str(index), Xmin=1100,
Xmax=19990, OutputWorkspace=str(i)+ '_' + str(index))
DeleteWorkspace(str(i))
#Some constraints included to precent divergence
fit_output = Fit(Function='name=Gaussian,Height=19.2327,\
PeakCentre=4843.8,Sigma=1532.64,\
constraints=(4600<PeakCentre<5200,1100<Sigma<1900);\
name=FlatBackground,A0=16.6099,ties=(A0=16.6099)',
InputWorkspace=str(i)+ '_' + str(index),
MaxIterations=1000, CreateOutput=True,
Output=str(i)+ '_' + str(index) + '_fit',
OutputCompositeMembers=True,
StartX=3800, EndX=6850, Normalise=True)
paramTable = fit_output.OutputParameters
# This catches some fits where the fit constraints are ignored,
# allowing the peak to fall far outside the nominal range
if paramTable.column(1)[1] < 4600.0 or paramTable.column(1)[1] > 5200.0:
DeleteWorkspace(str(i)+'_0_fit_Parameters')
DeleteWorkspace(str(i)+'_0_fit_Workspace')
DeleteWorkspace(str(i)+'_0')
DeleteWorkspace(str(i)+'_0_fit_NormalisedCovarianceMatrix')
reject.append(str(i))
continue
else:
uAmps.append(pcharge)
peak_centres.append(paramTable.column(1)[1])
peak_centres_error.append(paramTable.column(2)[1])
peak_intensity.append(paramTable.column(1)[0])
peak_intensity_error.append(paramTable.column(2)[0])
RunNo.append(str(i))
DeleteWorkspace(str(i)+'_0')
DeleteWorkspace(str(i)+'_0_fit_Parameters')
DeleteWorkspace(str(i)+'_0_fit_Workspace')
DeleteWorkspace(str(i)+'_0_fit_NormalisedCovarianceMatrix')
combined_data=np.column_stack((RunNo, uAmps, peak_intensity,
peak_intensity_error, peak_centres, peak_centres_error))
np.savetxt(Path2Save+'\peak_centres_'+cycle+'.csv',
combined_data, delimiter=", ", fmt='% s',)The FIA project can provide infrastructure for running the script out fetching the output to ingest into the data lake.
The performance of the TS1 methane moderator is currently tracked by fitting peaks from the incident monitor on PEARL. This analysis is run manually within the Mantid software and involves maintaining lists of runs and cycle names to feed into the script, along with executing the actual code.
This procedure can be automated and the output data should be ingested into the data lakehouse to enabled tools such as Superset to view the corresponding plots.
The basics of the analysis amount to the following script:
Script
The FIA project can provide infrastructure for running the script out fetching the output to ingest into the data lake.