PPdisk simdata (CASA 3.3)

From CASA Guides
Revision as of 18:57, 29 September 2011 by Ahale (talk | contribs)
Jump to navigationJump to search

Simulating Observations in CASA

Note: This is a work in progress.

Old version: PPdisk simdata (CASA 3.2).

To create a script of the Python code on this page see Extracting scripts from these tutorials.

Protoplanetary disk

  • Simdata version for CASA 3.2


Explanation of the script

Set simdata as current task and reset all parameters
 default("simdata")   
Specify sky model image
 modelimage         =  "input50pc_672GHz.fits"  
Image coordinate system can be verified
 imhead("input50pc_672GHz.fits")  
Image center can be identified
 # ia.open("input50pc_672GHz.fits")
 # ia.shape()
 # [257L, 257L, 1L, 1L]
 # ia.toworld([128.5,128.5])
 # {'numeric': array([  4.71239120e+00,  -4.01423802e-01,   1.00000000e+00,  6.72000001e+11])}
 # qa.formxxx("4.71239120rad",format='hms',prec=5)
 # '18:00:00.03052'
 # qa.formxxx("-0.401423802rad",format='dms',prec=5)
 # '-022.59.59.602743'
 # ia.done()
Brightness scale can be viewed with 'imstat' task
imstat("input50pc_672GHz.fits")
# ...
#  'max': array([  6.52469971e-05]),
# ...
# that's 0.0652 mJy/pixel.   
Let's call our project psim2
project            =  "psim2"
We'll leave the sky model the way it is: simdata will create psim2.skymodel CASA image since this model is a fits file, and most but not all of CASA routines can operate directly on fits
modifymodel        =  False  
skymodel           =  "input50pc_672GHz.fits"   
We need to decide where to point the telescope. The image is 2/3 arcsec in size, so we only need one pointing. We could put that in a text file ourself, or let simdata create the ascii pointing file for us.
setpointings       =  True
direction          =  "J2000 18h00m00.031s -22d59m59.6s"
mapsize            =  "0.76arcsec"  
The default pointingspacing is fine: we'll only fit one pointing in the small mapsize the default calculation maptype hexagonal is ok too since only one will fit anyway.
We do want to calculate visibilities in a measurement set: let's do a 20 min snapshot observation using out20 configuration:
predict            =  True
totaltime          =  "1200s"
Use appropriate antenna configurations based on desired angular resolution (configuration 20 is the largest "compact" configuration)
repodir=os.getenv("CASAPATH").split(' ')[0]
antennalist        =  repodir+"/data/alma/simmos/alma.out20.cfg"
Deconvolve the visibilities back into an image
image              =  True
vis                =  "$project.ms"
imsize             =  [192, 192]  
Specify number of iteration of cleaning task with proper threshold and weighting
niter              =  10000
threshold          =  "1e-7Jy"
weighting          =  "natural"    
We'd like to calculate a difference and fidelity image, and see some diagnostics:
analyze            =  True  
And see the array but not the UV coverage:
showarray          =  True
showuv             =  False  
Plot both to the screen and the png files with lots of messages:
graphics           =  "both"
verbose            =  True
overwrite = True

Run simdata

 CASA<> go simdata


  • Output results:


Input:
Predict:
Image:
Analyze:

Simulating Observations in CASA