Logged on 09/29/2009 10:28:18 PM
In the exercise we edited the wsrt_beams.py script to make beam shape irregular, in this example we also introduce pointing errors. Here are the inputs that were given to the script, note that we have options for the WSRT beam that are automatically displayed based on the information found in the TDL script example-sim.py, in particular by adding "wsrt_beams" to the list of sky_jones options -meqmaker.add_sky_jones('E','beam',[analytic_beams,sarod_cs1_beams], +meqmaker.add_sky_jones('E','beam',[analytic_beams,wsrt_beams,sarod_cs1_beams], pointing=oms_pointing_errors); and by adding the options as well as default values with TDLOption in the wsrt_beams.py script. _wsrt_option_menu = TDLCompileMenu('WSRT beam model options', - TDLOption('wsrt_beam_size_factor',"Beam size factor",[68.],more=float) + TDLOption('wsrt_beam_size_factor',"Beam size factor",[68.],more=float), + TDLOption('beam_ellipticity',"Beam ellipticity",[.01],more=float)
Logged on 09/29/2009 10:42:02 PM
Hmm, look what we've done!
We started with the measurement set WSRT_le.MS and generated new data from it, just see our pretty inconsistent wavy lines in our inspector!
Logged on 09/29/2009 10:52:20 PM
Cool we have an image or do we?
alas there appears to be 9 point source and they appear in a special pattern, pour quoi? Is there something wrong with the observation, perhaps the equipment, or perhaps two many alcoholic beverages?
No, no, it is the software! Yes, we successfully added affects into the image due issues with beam patterns being slightly different in different polarizations.
This is not a problem but a success in modelling the affects of beam patterns in an image.
WSRT_le.MS.CORRECTED_DATA.channel.64ch.fits (header) 64x1x512x512 FITS cube, 4 planes are given below. | ||||||||||||||
Image plane #0.
|
||||||||||||||
Image plane #1.
|
||||||||||||||
Image plane #2.
|
||||||||||||||
Image plane #3.
|
Logged on 09/29/2009 10:57:08 PM
Let's see what clean does -- will it remove the beam pattern effects or not?
Hmm, looks like not! We now have 9 very wonderful point sources now, some a little brighter then others but there are definitely 9.
Thus, we must do something with our data to get rid of this effect before applying clean.