MIR Tutorial (Charlie Qi 02/01/2012)
		 

1st half:

1. Tips

a. take notes

b. save and restore datasets

c. check the data filter

d. type 'retall' if anything crashes

2. General procedures -- Determine corrections vs freq,time and apply.

a. data inspection

b. tsys correction

c. bandpass cal

d. gain cal

e. data output: MIRIAD or FITS 

------------details---------------------

3. MIR setup and MIRIAD output setup

4. Step by step

a. data inspection:
   - Data structure: headers (in, bl, sp) and their corresponding pointers.
     * help, in,/str to get the variable names
     * blcd vs baseline; sb vs sideband; rec vs rx; band vs band; pol 
       tssb: tsys; integ: scan seconds
       wt: weighting; nch: channel numbers; prbl: projected baseline (m)
       ampave: averaged amplitude; phaave: averaged phase
     * Filtering: use wrappers vs function dat_filter; stackable
                  before reset
     * mir_save,/new: to create new subset.
   - plot_continuum 
     * c1 pseudo continuum band. can be generated by uti_avgband
     * uti_avgband uses the center 82 MHz of each 104 MHz chunk. 
   - plot_spectra
     * plot_spectra, intavg=10
     * select,/p,/re,source='IMLUP',sideband='u'
       plot_spectra,x='fsky',frame_v='sb',color='band', preavg=2
   - plot_var (take note on the IF1 and IF2 tsys)
     * plot_var,x='prbl',y='ampave',frame_var='sb',color='blcd'
     * plot_var,y='ampave',x='el',frame_var='blcd,sb',color='source'
     * plot_var,x='prbl',y='ampave',frame_v='sb',color_v='source'

b. tsys correction:
   - tsys replacement

c. bandpass cal. 
   - check atmospheric absorption ( e.g. at 358 ghz)
   - calibrator selection (compiled by Mark Gurwell)
      3C111=0418+380, BLLac=2202+422 have very deep, narrow CO, other 
      lines, other sources of concern would be 0359+509, 0530+135, 
      NRAO530=1733-130, and even 3C454.3 (though that is not as deep).

      For solar system sources, Neptune has broad CO absorption plus 
      an ~80 MHz wide emission peak (so bad there) as well as HCN 
      emission over about 100 MHz.

      Titan has lines all over the place so avoid.

      Mars, Venus have ~200 MHz CO absorption lines that can be up to 
      40% of the continuum in depth.  Isotopes can exhibit ~5%
      absorption as well.

      Uranus has no known spectral lines.  Callisto and Ganymede are 
      also very good (solid bodies, no appreciable atmosphere).  
      Same with Mercury if out of sun avoidance zone.

      Jupiter and Saturn may be ok, but when they are highly resolved 
      may not be sufficient because nulls can happen within the 4 GHz 
      IF and that will screw things up. Saturn 279 GHz PH3 line ?

   - recommend baseline-based amplitude bandpass cal, see log 15345
   - separate phase and amplitude bandpass cal. 
     the usual procedures:
         antenna-based phase bandpass cal
	 regenerate continuum with uti_avgband
	 baseline-based amplitude bandpass cal
   - keywords: preavg, ntrim, smoothing

d  gain cal
   - get flux information first.
   - keyword preavg

e  output
   - idl2miriad, libfile='filelocation' ...

--------------
2nd half

1. baseline correction.
 a. see reference page in SMA Observer Center.
 b. check antennas file.

2  flux calibration and measurement

 a. source selection (compiled by Mark again)
    Best - Titan
    Good - Callisto, Uranus*, Neptune* (if 12+GHz away from CO lines)
    Almost as good - Mars*, Ganymede
    Not Good - Saturn, Venus, Jupiter, Pluto
    Here, * means beware of overresolving the source!

 b. IDL> select,/p,/re, band='c1'
    IDL> gain_cal,cal_type='pha',/connect,/non_point,tel_bsl='telescope',
                    refant=?
          - type "all yes" when it asks for calibrator names and fluxes
          - then type "yes" to apply the solution
    IDL> select,/p,/re, band='c1'
    IDL> gain_cal,x='hours',poly=0,cal_type='amp',/preavg,/non_point,
                   tel_bsl='telescope',refant=?
          - type "all no" and then type "callisto yes"
          - then type "yes" to apply the solution.
    IDL> select,/p,/re, band='c1'
    IDL> flux_measure
          - type 'V' for vector average

 c. compare with Mark's table.  
    - plot_var,x='prbl',y='ampave',frame_v='sb',color_v='source'


3. DBA track, if1/if2 phase offsets. 
   - uti_difif to check the difference of if1/if2
     plot_continuum,y='difif',/preavg

4. new tsys replacement program (more recent data with tsys_read)
   - select,/p,/re
     uti_tsys,/replace

5. line identification. -- mess around sky freq header !
   - uti_restfreq

6. multiple sources dopplerTracking
   - uti_doppler_fix
 
7. data fix (see MIR cookbook webpage) 

 a. "This dualrx track might have a wrong data structure .
      Please report this track directory to cqi@cfa."
      - IDL> bl.irec=1

 b. idl2miriad library location.