NAME

ripple - Compute the GHRS ripple function coefficients.

USAGE

ripple input1 input2 output1 output2

DESCRIPTION

This task computes the correction coefficients to a theoretical Echelle ripple equation for GHRS. Using a table containing the ripple profile of one or more spectral orders--tabulated as a function of carrousel position and spectral order--a non-linear least squares fit is performed to fit a, b, and c (which are computed using the Levenberg-Marquardt method) in the following equation:

		R = c * N * sinc( a*X + b)
 
	where:
	        pi * m * cos(theta + beta + delta) * sin(theta)
	    X = -----------------------------------------------
			       sin(theta + beta)

	        cos(theta + beta + delta)
	    N = -------------------------
	        cos(theta + beta - delta)  
 
	    beta  - is the Echelle blaze angle,
	    delta - is the half-angle between the collimator and
			camera position,
	    theta - is the scan angle computed by:
 
			   (R0 - carrousel position)
		   theta = -------------------------  - beta
				   182.0444
 
	    R0     - is the carrousel position at the Echelle blaze.

PARAMETERS

input[file name]

The name of the table containing the ripple profile tabulated as a function of the spectral order and carrousel position. This table is usually produced by the abssenz task and must contain the following columns:

     'ORDER'  - Spectral order number (integer).
     'CARPOS' - Carrousel position (integer).
     'VALUE'  - Ripple profile value (double).

input[file name]

The name of the table containing the grating parameters. This table must have one row with the columns (angles are in degrees) BETA, DELTA, and R0.

output= "ripcoef" [file name]

Output table containing the fitted coefficients required by the Routine Science Data Processing (RSDP) system. RSDP requires tabulated values for each order and multiple carrousel positions. RSDP will interpolate in the carrousel position, if necessary, to compute values of a and b. The output table therefore has rows for the minimum and the maximum carrousel position calibrated. For each carrousel position, a row is written for every order calibrated. The output table will contain the following columns:

     'GRATING' - Grating mode (char*5).
     'CARPOS'  - Carrousel position (integer).
     'SPORDER' - Spectral order (integer).
     'A'       - Coefficient a (double).
     'B'       - Coefficient b (double).

output= "ripplefit" [file name]

Table containing results of the fit for each input data point. This table has the same columns as input1, but, has an additional column called FIT which contains the fitted values.

(a = 1.0) [real, min=0.1, max=2.0]

Initial guess for the coefficient a.

(b = 0.0) [real]

Initial guess for the coefficient b.

(niter = 20) [integer, min=1, max=500]

Number of iterations in the non-linear least squares routine. Normally 20 should be more than sufficient.

(carpos= 0) [integer, min=0, max=65536]

The minimum carrousel position for which the calibration is to be used. If carpos1 = 0, then the lowest carrousel position in the input data will be used.

(carpos= 0) [integer, min=0, max=65536]

The maximum carrousel position for which the calibration is to be used. If carpos2 = 0, the highest input carrousel position is used.

(m= 0) [integer, min=0, max=100]

The lowest spectral order for which the calibration is to be used. If m1 = 0, the minimum input spectral order is used.

(m= 0) [integer, min=0, max=100]

The highest spectral order for which the calibration is to be used. If m2 = 0, the maximum input spectral order is used.

(aout) [real]

Output parameter giving the computed value of the coefficient a.

(bout) [real]

Output parameter giving the computed value of the coefficient b.

EXAMPLES

1. Generate the coefficients for the Echelle ripple using the input profile stored in table riptab and the grating coefficients in table gcoef.

  cl> ripple riptab gcoef

2. Do 10 more iterations using the results of a and b as initial guesses.

  cl> ripple.a = ripple.aout
  cl> ripple.b = ripple.bout
  cl> ripple riptab gcoef niter=10

BUGS

REFERENCES

HELP

For assistance using this or any other tasks, please contact help@stsci.edu or call the help desk at 410-338-1082.

SEE ALSO

centerflux, abssenz