CutMD v1¶

See Also¶

SliceMDHisto, ProjectMD, SliceMD, BinMD

MDEventWorkspaces¶

For input of type MDEventWorkspace the algorithm uses BinMD v1 or SliceMD v1 to achieve the binning of the data. The choice of child algorithm used for slicing in this case is controlled by the NoPix option.

MDHistoWorkspaces¶

If the input is an MDHistoWorkspace BinMD v1 and SliceMD v1 are not made available as they needto get hold of the original MDEvents associated with an MDEventWorkspace in order to perform the rebinning. As this information is missing from the MDHistoWorkspace images, those operations are forbidden. Instead, a limited subset of the operations are allowed, and are performed via IntegrateMDHistoWorkspace v1. In this case, the Projection and NoPix properties are ignored. See IntegrateMDHistoWorkspace v1 for how the binning parameters are used.

Projection Binning¶

The ‘PnBin’ property, where n is between 1 and 5, is used to specify the binning for the nth dimension of the output workspace. The dimension will be truncated to have extents ‘minimum’ and ‘maximum’, with ‘stepsize’ specifying the size of the bins inbetween. ‘(maximum - minimum)/stepsize’ is rounded down to produce an integer number, greater than or equal to 1, of equally-sized bins between ‘minimum’ and ‘maximum’.

Note that if the output workspace is an MDEventWorkspace (NoPix=False), these properties define the top-level box structure of the workspace. If the MaxRecursionDepth property is set to higher than its default of 1 then if many events fall within a single box it may be split further, see the documentation for MDEventWorkspace.

The PnBin parameters must match one of these three formats:

For ease of use, when using the python interface only, the ‘PBins’ keyword can be used in place of separate PnBins properties. PBins accepts a tuple, or list, of PnBins parameters. The position in the list determines the dimension it corresponds to. See the Usage examples below.

Creating Projections¶

Projections are used by CutMD to transform the multidimensional data prior to cutting it. Projections are provided to CutMD in the form of a TableWorkspace. The format of these workspaces is as follows:

A projection table should have three rows: one for u, one for v, and one for w.

There is a helper class called Projection that can be used to construct these projection tables for you automatically. For example:

from mantid.api import Projection
# Create an identity projection
proj_id = Projection([1,0,0], [0,1,0], [0,0,1])

# Automatically infer third dimension as being orthogonal to the first two
proj_rot = Projection([1,1,0], [1,-1,0])

# Set other properties
proj_prop = Projection()
proj_prop.setOffset(0, 100) # Set u offset to 100
proj_prop.setOffset(1, -5.0) # Set v offset to -5
proj_prop.setType(1, 'r') # Set v type to be RLU
proj_prop.setType(2, 'a') # Set w type to be inverse angstroms

#Create table workspaces from these projections
ws_id = proj_id.createWorkspace() # Named ws_id
proj_rot.createWorkspace(OutputWorkspace="ws_rot") # Name ws_rot

When calling createWorkspace inside of algorithms like CutMD, the OutputWorkspace name must be provided, or createWorkspace will not know what to call the created workspace:

#Good:
CutMD(..., Projection=proj.createWorkspace(OutputWorkspace='proj_ws'), ...)

#Bad:
CutMD(..., Projection=proj.createWorkspace(), ...)

Workflow¶