Tests for working mstransform routine

[iancze]

As a validation workflow, we want to compare two paths to (hopefully) the same science product

• take an MS, input a reasonable tclean command, and then get out a science quality image cube
• apply mstransform or our own routines to barycorr and regrid the visibilities, then use tclean with cubedata to get a science quality image cube

Both cubes should be identical, or nearly so.