.. _01_pim_sol:

============================
Python in Mantid: Solution 1
============================

All the data for these solutions can be found in the TrainingCourseData on the Downloads page.

A - Using ISIS Data
===================

.. code-block:: python

    from mantid.simpleapi import *

    # The input data set
    inputData = "HRP39182"
    # Load the file
    Load(Filename=inputData+".RAW",OutputWorkspace=inputData,Cache="If Slow")

    # First do the analysis without prompt pulse removal so that we can compare the difference

    cal_file = "hrpd_new_072_01_corr.cal"
    ApplyDiffCal(InstrumentWorkspace=inputData, CalibrationFile=cal_file)
    ConvertUnits(InputWorkspace=inputData, OutputWorkspace="aligned-withpulse", Target="dSpacing")
    # Focus the data
    DiffractionFocussing(InputWorkspace="aligned-withpulse",OutputWorkspace="focussed-withpulse",GroupingFileName=cal_file)

    # Remove the prompt pulse, which occurs at at 20,000 microsecond intervals. The bin width comes from a quick look at the data
    for i in range(0,5):
      min = 19990 + (i*20000)
      max = 20040 + (i*20000)
      MaskBins(InputWorkspace=inputData,OutputWorkspace=inputData,XMin=min,XMax=max)

    ApplyDiffCal(InstrumentWorkspace=inputData, CalibrationFile=cal_file)
    ConvertUnits(InputWorkspace=inputData, OutputWorkspace="aligned-withoutpulse", Target="dSpacing")
    # Focus the data
    DiffractionFocussing(InputWorkspace="aligned-withoutpulse",OutputWorkspace="focussed-withoutpulse",GroupingFileName=cal_file)
    # Subract the processed data with and without pulse from eachother
    Minus(LHSWorkspace="focussed-withpulse", RHSWorkspace="focussed-withoutpulse", OutputWorkspace="difference")

    # Now plot a focussed spectrum with and without prompt peak removal so that you can see the difference
    plotSpectrum(["focussed-withoutpulse","difference", "focussed-withpulse"],0)


.. plot::

    from mantid.simpleapi import *
    from mantid.plots._compatability import plotSpectrum

    # The input data set
    inputData = "HRP39182"
    # Load the file
    Load(Filename=inputData+".RAW",OutputWorkspace=inputData,Cache="If Slow")

    # First do the analysis without prompt pulse removal so that we can compare the difference
    # Align the detectors (incoporates unit conversion to d-Spacing)
    cal_file = "hrpd_new_072_01_corr.cal"
    ApplyDiffCal(InstrumentWorkspace=inputData, CalibrationFile=cal_file)
    ConvertUnits(InputWorkspace=inputData, OutputWorkspace="aligned-withpulse", Target="dSpacing")
    # Focus the data
    DiffractionFocussing(InputWorkspace="aligned-withpulse",OutputWorkspace="focussed-withpulse",GroupingFileName=cal_file)

    # Remove the prompt pulse, which occurs at at 20,000 microsecond intervals. The bin width comes from a quick look at the data
    for i in range(0,5):
      min = 19990 + (i*20000)
      max = 20040 + (i*20000)
      MaskBins(InputWorkspace=inputData,OutputWorkspace=inputData,XMin=min,XMax=max)

    # Align the detectors (on the data with the pulse removed incoporates unit conversion to d-Spacing)
    ApplyDiffCal(InstrumentWorkspace=inputData, CalibrationFile=cal_file)
    ConvertUnits(InputWorkspace=inputData, OutputWorkspace="aligned-withoutpulse", Target="dSpacing")
    # Focus the data
    DiffractionFocussing(InputWorkspace="aligned-withoutpulse",OutputWorkspace="focussed-withoutpulse",GroupingFileName=cal_file)
    # Subract the processed data with and without pulse from eachother
    Subtract(LHSWorkspace="focussed-withpulse", RHSWorkspace="focussed-withoutpulse", OutputWorkspace="difference")

    # Now plot a focussed spectrum with and without prompt peak removal so that you can see the difference
    plotSpectrum(["focussed-withoutpulse","difference", "focussed-withpulse"],0)


B - SNS Data
============

Right-click in the Messages toolbox and check that the log level is set to "Notice" (or lower, such as "Information" and "Debug").

.. code-block:: python

    from mantid.simpleapi import *

    run = Load('PG3_4871_event')
    nevents = run.getNumberEvents()
    logger.notice('Number of Events Before = {}'.format(str(nevents)))
    filtered = FilterBadPulses(run, LowerCutoff=99.5)
    ApplyDiffCal(InstrumentWorkspace=filtered, CalibrationFile='PG3_golden.cal')
    aligned = ConvertUnits(InputWorkspace=filtered, Target="dSpacing")

    rebinned = Rebin(InputWorkspace=aligned, Params=[1.4,-0.0004, 8])
    focused = DiffractionFocussing(InputWorkspace=rebinned, GroupingFileName='PG3_golden.cal')
    compressed = CompressEvents(InputWorkspace=focused)
    nevents = compressed.getNumberEvents()
    logger.notice('Number of Events After = {}'.format(str(nevents)))

Output:

.. code-block:: none

    Number of Events Before = 22065736
    Number of Events After = 555305

C - ILL Data
============

For point 5., you are expected to open the docs page for :ref:`algm-ConvertUnits` and notice that `EFixed` should be set to the Ei value (Found in Show Sample Logs). This docs page also mentions that for Ei, the parameter EMode='Direct' is required. These parameters are required by the target of the unit conversion - DeltaE.

.. code-block:: python

    from mantid.simpleapi import *

    # You can load each file individually, e.g.
    # 164198 = Load(Filename=164198)
    # ... for each file

    # Or you can Load data in a loop, but you may need to import the workspaces to Python
    for i in range(164198,164201):
        print( '{}.nxs'.format(i))
        Load(Filename = '{}.nxs'.format(i), OutputWorkspace = str(i))

    mtd.importAll()

    data_merged = MergeRuns([164198,164199,164200])
    bad_spectra = [1,2,3,4,5,6,11,14,30,69,90,93,95,97,175,184,190,215,216,217,251,252,253,255,289,317,335,337]
    MaskDetectors(Workspace = data_merged, SpectraList = bad_spectra)
    scaled = MultiplyRange(data_merged, Factor = 0.95)
    ws = ConvertUnits(scaled, Target = 'DeltaE', EFixed = 4.7728189558864003, EMode = 'Direct') # See explanation above
    wsCorrected = DetectorEfficiencyCorUser(ws)

    print("The corrected value in spectrum with ws index {}, bin {} is {:.2f} compared to {:.2f}".format(6,4,wsCorrected.readY(6)[4],ws.readY(6)[4]))

Output:

.. code-block:: none

    The corrected value in spectrum number 6, bin 4 is 278.85 compared to 0.95