Using Tpoint

The Tpoint (TM) program by Patrick Wallace is a tricky beast. After much trial and error, I have come up with a way of getting a pointing model that is quite usable, especially when the data from the whole sky is available. I describe here my ad-hoc method of using Tpoint to get a satisfactory pointing model. If you make a mistake, the easiest thing to do is quit Tpoint and restart the fitting procedure. This section is meant to be used in conjunction with Section 5.2.4 or Section 5.4. What is done with the output model depends on whether or not the telescope has been polar aligned.

1. Star Tpoint In the directory with the .dat file output by rstarwrite

   $ tpoint

   ...

   TPOINT ready for use:  type HELP for assistance, END to quit.

   *

2. Load in the Tpoint data file.

   * indat file.dat

   ROTSE-III Prototype

   :EQUAT

   :NODA

   :ALLSKY

   35 52 8.4  2001 5 30
   

   14 16  0.00  19 12  0.00 20 20 30.00 164  4   0.00 13 40.87

   ...

   A list of all the stars logged should be printed to the screen.

3. Set Tpoint to adjust star positions to fit telescope positions. This is the modelling that schierd uses.

   * adjust s

   Model will adjust stars to fit telescope.

4. First, have Tpoint fit to the index offsets for the ra and dec axes. The ``use'' feature tells the Tpoint to fit the stars to these particular variables.

   * use ih id

5. Next, do the fit to these parameters. The fit command should be run several times, until the errors reported stop shrinking.

   * fit

          coeff       change     value    sigma

     1     IH         -0.000-319288.44  852.791

     2     ID         -0.000 -12374.29  654.834

   Sky RMS =1464.25

   Popn SD =1890.34

   When the ``change'' column sits at 0, you can move on.

6. Fix the index offsets with fix, and now have Tpoint use the collimation and polar alignment errors for its next fit.

   * fix ih id

   * use np ch me ma

7. Fit with the fit command several more times until the change goes to 0, as before.

8. Use the whole suite of parameters, by unfixing IH and ID and fit several more times until Tpoint settles on a solution.

   * use ih id

   * fit

   ...

9. Save the model to disk and exit.

   * outmod modelfile.dat

   * end

10. The model file can now be viewed with a textfile viewer to extract notable parameters, or can be used verbatim by schierd.conf as described in Section 3.8. The model file should look something like the following:
    

    ROTSE-III Prototype

    S  1862248.2854    0.000   0.0000

      IH      -320481.5133   114.58002

      ID      -15074.2754    15.45639

      NP       -284.0514    24.71350

      CH       -356.0703   109.36628

      ME      -2233.0871    18.88705

      MA       -728.2158     9.84367

    END

    The column to the right of the fit parameter names contains the fit values, in arcseconds. The third column contains the rms error in the fit for that parameter.