next up previous contents index
Next: Comparison and practical advices Up: Implementation and typical use Previous: Implementation   Contents   Index

Typical deconvolution session

       1 read beam demo
       2 read dirty demo
       3 input clean
       4 hogbom /flux 0 1
       5 show residual
       6 show clean
       7 write clean demo
       8 let name demo
       9 go noise
      10 let ares 0.5*noise
      11 input clean
      12 hogbom /flux 0 1
      13 let niter 2000
      14 input clean
      15 hogbom /flux 0 1
      16 show residual
      17 show clean
      18 for iplane 1 to 10
      19    show clean iplane
      20    support
      21    hogbom iplane /flux 0 1
      22    write support "demo-"'iplane'
      23 next iplane
      24 show residual
      25 show clean
      26 write residual demo
      27 write clean demo
      28 write cct demo
      29 let name demo
      30 let type lmv-clean
      31 go cct
      32 go view
Comments:
Steps 1-2
Read dirty beam and dirty image from the demo.beam and demo.lmv files. Those steps are not needed if the dirty beam and image are already stored in the internal buffer, i.e. if you have imaged the $uv$ table just before in the same MAPPING session.
Steps 3-6
Print the current state of the control parameters, deconvolve the dirty image using the HOGBOM algorithm (step 3) and look at the results (residual and clean images). The /flux 0 1 option pop-up the visualization of the cumulative flux deconvolved as the clean components are found.
Steps 8-12
Estimate the empirical noise through the GO NOISE command after this first deconvolution and set the ares stopping criterion accordingly. Check that the new value of ares has been correctly set (step 11) and restart deconvolution.
Steps 13-17
Increase the number of clean components as the previous deconvolution stopped before that the residual image reached the ares value. Restart deconvolution and look at results.
Steps 18-25
Tries to improve deconvolution by definition of a support per plane and deconvolve this plane accordingly. The support is store in a file for memory. The deconvolution results are then displayed.
Steps 26-28
Write residual image, clean image and clean component list in demo.lmv-res, demo.lmv-clean and demo.cct files for memory.
Steps 29-32
Visualize the cumulative flux as a function of the clean component number and visualize the clean spectra cube in an interactive way.
Typical deconvolution session using other CLEAN algorithm would look very similar. The main difference would be the possible tuning of other control parameters. A deconvolution session using MX would start differently as the imaging and deconvolution are done in the same step:
       1 read uv demo
       2 input uv_map
       3 input clean
       4 mx /flux 0 1
       5 show residual
       6 show clean
       7 write beam demo
       8 write dirty demo
       9 write clean demo
      10 write residual demo
      11 write cct demo
Comments:
Step 1
Read the demo.uvt $uv$ table in an internal buffer.
Steps 2 and 3
Check current state of the variables that control the imaging and deconvolution.
Steps 4-6
Deconvolve and look at the results.
Steps 7-11
Write all the internal buffers for memory.
All the tuning of the typical imaging and deconvolution sessions could be fitted in this MX session although they are not repeated here.


next up previous contents index
Next: Comparison and practical advices Up: Implementation and typical use Previous: Implementation   Contents   Index
Gildas manager 2014-07-01