next up previous contents
Next: 1.7 Separation of spectral Up: 1 The MSTransform framework Previous: 1.5 Channel averaging   Contents

1.6 Reference frame transformation

The parameter regridms in mstransform can do what task cvel does, such as regrid an MS to a new spectral window, channel structure or frame and it is much faster in all these transformations. The option regridms in mstransform differs from task cvel in a few cases. Task cvel has a parameter passall to copy or not the non-selected spws into the output MS. In mstransform we only consider the passall=False, meaning that we only consider the selected spws of the MS. Task cvel always combines spws when changing the reference frame, while mstransform can do both cases. The combination of spectral windows in mstransform is controlled by the independent parameter combinespws.

Mstransform will only shift spws with channel widths of the same sign in a single operation. If you are regridding spws with mixed positive and negative channel widths, you should run this task separated for each group of spws. You can verify the channel widths for your MS using listobs for example, by looking at the SPW table, column ChanWid. Whenever the width parameter is > 2, a pre-averaging is done.

Gaps between spectral windows are padded with interpolated data, therefore it is not advisable to combine upper and lower sidebands with a large gap as this will bloat the data set.

    mstransform('uid.ms', outputvis='myout.ms', spw='1~3', regridms=True, nchan=10, outframe='LSRK')


next up previous contents
Next: 1.7 Separation of spectral Up: 1 The MSTransform framework Previous: 1.5 Channel averaging   Contents
Sandra Castro 2014-12-16