[daip] Maxfit not converging

[Eric Greisen]

Katherine Blundell wrote:
>
>> You have not told me the actual beam width and the signal to noise
>> This may be as good as it gets
> 
> Sorry I forgot to mention this:
> Conv size= 0.00973 X 0.00497   Position angle=  -2.40
> cellsize = 0.0015

> It gives quite a different answer - I've attached the JMFIT below, to be 
> compared with the MAXFIT output of:
> 
> maxfit
> AIPS 1: Map pixel position:  1048.22  956.22
> 
> Thanks for your help with this - we've been pretty flummoxed with it all.  Is there any chance of slightly amending JMFIT so that it outputs the difference between the fitted peak and the nearest centre of pixel to give rashift and decshift in the right directions?

> 
> asosx1> JMFIT1: Task JMFIT  (release of 31DEC12) begins
> asosx1> JMFIT1: Fit to histogram gives RMS = 1.431E-04
> asosx1> JMFIT1: Using Clean beam   0.00975   0.00497   -2.69 from CG table
> asosx1> JMFIT1: Model fit to SS433_C0    .ICL001.  16
> asosx1> JMFIT1: Source= SS433_C0 in units JY/BEAM
> asosx1> JMFIT1: X-ref pix=1024.0  RA 19 11 49.5666000  pix sep (Asec) 
> -0.0015
> asosx1> JMFIT1: Y-ref pix=1025.0  DEC 04 58 57.873000  pix sep (Asec) 
>  0.0015
> asosx1> JMFIT1: Axes 3-7 pixels =      1      1      1      1      1
> asosx1> JMFIT1: Axis 3 : IPOL
> asosx1> JMFIT1: Axis 4 : 1666.428 MHZ
> asosx1> JMFIT1:  - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> asosx1> JMFIT1: ******** Input Model 
> ****************************************
> asosx1> JMFIT1: Component=  1  Gaussian
> asosx1> JMFIT1:       Peak= 3.0135E-02    JY/BEAM
> asosx1> JMFIT1:       Xpos= 1048.00      Ypos=  956.00     pixels
> asosx1> JMFIT1:       Size=   6.50    x   3.31    Pixels at pa  -2.69    deg
> asosx1> JMFIT1:  - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> asosx1> JMFIT1:       Beam=   6.50    X   3.31    pixels in PA   -2.7    deg
> asosx1> JMFIT1:
> asosx1> JMFIT1: Initial guess RMS  1.5170E-03 in10000 usable pixels
> asosx1> JMFIT1: Fitting to10000 data points for  6 parameters
> asosx1> JMFIT1: Initial Chi-squared, R.M.S., and gradient norm:
> asosx1> JMFIT1:               2.30131E-02    9.14568E-06    8.84970E-01
> asosx1> JMFIT1: ***  Convergence achieved.
> asosx1> JMFIT1: Solution RMS  1.0347E-03 in10000 usable pixels
> asosx1> JMFIT1: Post-fit Chi-squared, R.M.S., and gradient norm:
> asosx1> JMFIT1:               1.07061E-02    1.03501E-03    3.81988E-05
> asosx1> JMFIT1: Writing MF file version   1 attached to input image
> asosx1> JMFIT1:
> asosx1> JMFIT1: ********* Solution from JMFIT 
> *********************************
> asosx1> JMFIT1:
> asosx1> JMFIT1: Component   1-Gaussian
> asosx1> JMFIT1:   Peak intensity    = 2.5168E-02 +/-  1.31E-04 JY/BEAM
> asosx1> JMFIT1:   Integral intensity= 1.0566E-01 +/-  6.67E-04 JANSKYS
> asosx1> JMFIT1:   X-position        =   1047.572 +/-   0.0225 pixels
> asosx1> JMFIT1:   Y-position        =    956.036 +/-   0.0217 pixels
> asosx1> JMFIT1:                       RA 19 11 49.5642339  +/-  0.000002263
> asosx1> JMFIT1:                       DEC 04 58 57.769554  +/-   0.00003254
> asosx1> JMFIT1:   Major axis        =  12.029 +/-  0.0625 pixels
> asosx1> JMFIT1:   Minor axis        =   7.508 +/-  0.0390 pixels
> asosx1> JMFIT1:   Position angle    = 132.479 +/-  0.430 degrees
> asosx1> JMFIT1:   Major axis        = 0.0180440 +/- 0.0000937 asec
> asosx1> JMFIT1:   Minor axis        = 0.0112624 +/- 0.0000585 asec
> asosx1> JMFIT1:   Position angle    = 132.480 +/-  0.430 degrees
> asosx1> JMFIT1:   - - - - - - - - - - - - - - - - - - - - - - - - - - - 
> - - -
> asosx1> JMFIT1:                   Deconvolution of component in pixels
> asosx1> JMFIT1:                        Nominal     minimum     maximum
> asosx1> JMFIT1:     Major ax            10.987      10.909      11.064
> asosx1> JMFIT1:     Minor ax             5.212       5.132       5.292
> asosx1> JMFIT1:     Pos ang            122.709     122.114     123.308
> asosx1> JMFIT1:                   Deconvolution of component in asec
> asosx1> JMFIT1:                        Nominal     minimum     maximum
> asosx1> JMFIT1:     Major ax          0.016480    0.016364    0.016596
> asosx1> JMFIT1:     Minor ax          0.007819    0.007698    0.007938
> asosx1> JMFIT1:     Pos ang         122.709045  122.113564  123.307846
> asosx1> JMFIT1:   - - - - - - - - - - - - - - - - - - - - - - - - - - - 
> - - -
> asosx1> JMFIT1: returns adverbs to AIPS
> asosx1> JMFIT1: Appears to have ended successfully
> asosx1> JMFIT1: asosx120 31DEC12 TST: Cpu=      0.1  Real=    133  IO=   
>      33

I changed the output of MAXFIT to the adverb FSHIFT (SHIFT is used a 
variety of places and one does not want it changing for non-obvious 
reasons).  I have then added FSHIFT to IMFIT and JMFIT as an output 
adverb and is the printed display.  I hope this helps but sometimes 
things just are not well enough represented at this fine level by 
symmetric functions.

Cheers

Eric