Table of Contents
Name | Direction | Type | Default | Description |
---|---|---|---|---|
InputWorkspace | Input | Workspace | Mandatory | input workspace |
OutputWorkspace | Output | Workspace | Mandatory | output workspace |
XMin | Input | dbl list | minimum x values with NaN meaning no minimum | |
XMax | Input | dbl list | maximum x values with NaN meaning no maximum |
This is a workflow algorithm that crops each spectrum of a workspace independently. This is intended for workspaces with a relatively small number of spectra (e.g. <10), but places no restrictions on the input workspace.
The minimum and maximum values that are specified are interpreted as follows:
Note
To run these usage examples please first download the usage data, and add these to your path. In MantidPlot this is done using Manage User Directories.
This is an example of how CropWorkspaceRagged would be used near the end of a workflow to generate a real-space distribution of data after it had been reduced into a number of “banks” or “spectra.” As mentioned above, numpy.nan or math.nan can both be used.
from numpy import nan
NOM_91796 = LoadNexusProcessed(Filename='NOM_91796_banks.nxs')
CropWorkspaceRagged(InputWorkspace=NOM_91796, OutputWorkspace='cropped',
Xmin=[0.67, 1.20, 2.42, 3.70, 4.12, 0.39],
Xmax=[10.20, 20.8, nan, nan, nan, 9.35])
binning=(0.,.02,40.)
Rebin(InputWorkspace='cropped', OutputWorkspace='cropped',
Params=binning)
# put into a single spectrum and Fourier transform
SumSpectra(InputWorkspace='cropped', OutputWorkspace='FQ',
WeightedSum=True, RemoveSpecialValues=True)
PDFFourierTransform(InputWorkspace='FQ', OutputWorkspace='Gr',
InputSofQType='Q[S(Q)-1]', DeltaR=.02)
Categories: AlgorithmIndex | Transforms\Splitting | Workflow
Python: CropWorkspaceRagged.py (last modified: 2018-10-05)