autoMask

Overview

This utility is able to generate automatic pixel masks, which are applicable for certain sample environments. Currently it only works for the Paris-Edinburgh cells, which are commonly used on SNAP. It requires an input workspace that is an unfocussed workspace, containing data from a sample measured in a PE cell and which should normally be in units of wavelength (TOF is also usually OK). It then converts the data into a simple 2d image, by summing the events in each pixel. It then applies a thresholding algorithm (the “Li thresholding” algorithm from the scikit-image package) to identify masked pixels. Finally, the image mask is converted into a mantid pixel mask workspace that can then be used during data reduction.

Optional arguments

maskType

This is envisaged for future expansion of the mask provision. Currently, the default option “PE” for Paris-Edinburgh cells is the only supported mask type.

plotOn

Defaulting to True, this will pop up a 2d plot visually showing the mask that has been created. Changing the value of this parameter to False will prevent this image being shown.