2.Flagging

If you look closely at the previous plotms, there are some drop-outs. Plot TW Hya by itself, changing the x-axis to time and restricting the uv range to less than 50 m (where the source is brightest). Change the inputs in plotms or enter the following:

# In CASA
default(plotms)
vis='sis14_twhya_calibrated_flagged.ms' 
xaxis='time'
yaxis='amp' 
avgchannel='384'
avgtime='30s'
field='TW Hya'
uvrange='<50'
coloraxis='baseline'
plotms

Amplitude v. time for TW Hya on short baselines, colourised by baseline

Use select and locate to identify the bad data. Look in the logger. The fragment below shows the important details: In scan 16, all the bad data involve antenna DV19 and in scan 36, all the bad data involve DV20.

Scan=16 Field=TW Hya[5] Time=2012/11/19/08:08:24.7 BL=DV08@A021 & DV19@A008[14&21] 
Scan=16 Field=TW Hya[5] Time=2012/11/19/08:08:24.7 BL=DV19@A008 & DV22@A011[21&24] 
Scan=16 Field=TW Hya[5] Time=2012/11/19/08:08:24.7 BL=DV19@A008 & DV23@A007[21&25] 

Scan=36 Field=TW Hya[5] Time=2012/11/19/09:05:30.7 BL=DA48@A046 & DV20@A020[5&22] 
Scan=36 Field=TW Hya[5] Time=2012/11/19/09:05:30.7 BL=DV06@A037 & DV20@A020[12&22] 
Scan=36 Field=TW Hya[5] Time=2012/11/19/09:05:30.7 BL=DV08@A021 & DV20@A020[14&22] 

Use flagdata to flag the first set of bad data:

# In CASA
default(flagdata)
vis='sis14_twhya_calibrated_flagged.ms' 
antenna='DV19'
scan='16'
inp 
flagdata

You can see the flagging progress in the logger. Now change the parameters values to specify the other bad data noted above. If you also enter these (and further values you have to fill in) in twhya.py, you will have a script you can run later.

# In CASA
default(flagdata)
vis='sis14_twhya_calibrated_flagged.ms' 
antenna='**'                    # The other bad antenna 
scan='**'                       # The other bad scan 
inp 
flagdata

Check in plotms by ticking the reload button at the bottom and press plot. Then set the x-axis to channel and remove the uv range and channel averaging, but average all times, scans and baselines (see figure below). You can see a hint of a line but it so faint we can ignor it and average all channels for initial continuum imaging.


Spectrum of target data before self-cal, averaging all time including across scans and all baselines.

amsr@jb.man.ac.uk 15 Oct 2015