CD.solver: Measure central world coordinates and rotation angles
In esplint/IRACpm: Apply Distortion Correction to SPITZER IRAC Data

Description Usage Arguments Details Value Author(s) Examples

Measures central world coordinates and rotation angles for each frame in a data set. Allows options to specify which stars to use, select a lower SNR limit, have multiple core processing, provide independent star coordinate guesses, and provide independent central coordinates and rotation angle starting guesses.

CD.solver(data, realz = NA, cores = 4, SNR_cut = 20, 
star_id = NA, cda = NA)

CD.solver2(data, realz = NA, cores = 4, SNR_cut = 20, 
star_id = NA, cda = NA)

CD.solver3(data, realz = NA, cores = 4, SNR_cut = 20, 
star_id = NA, cda = NA)

CD.solver4(data, realz = NA, cores = 4, SNR_cut = 20, 
star_id = NA, cda = NA)

`data`	Dataset containing all the individual detections (i.e., Output of read.in.data).
`realz`	Optional initial guesses for world coordinates of stars. Should have same length as index$star_id
`cores`	Number of cores to run this function for multicore processing. Defaults to 4.
`SNR_cut`	Specify which detections should be ignored based on SNR lower limit. Defaults to SNR < 20.
`star_id`	Option to specify which stars to use in calculations. All other detections are ignored. Defaults to using all available stars.
`cda`	Optional initial guesses for individual image world coordinates and rotation. Should have same length as data$image_key$image_id

Since true coordinates of stars are not initially known, the function relies on an iterative approach. Unless cda is specified, the first guess is derived from the image headers recorded in data$image_key. Unless realz is specified, the estimate for star world coordinates comes from calc_all (We strongly recommend using 2MASS coordinates for realz). Function then iterates measuring image parameters using a least squares approach and recalculating star world coordinates. Outlier rejection is also applied twice. Uses internal functions: single_image_resid, coor_resid, and imageCD_solver, which are not designed for external computations. The time-dependent scale factor is estimated by linear interpolation between the known points already measured.

CD.solver2 does not revaluate the world coordinates but rather matches detections to realz after applying aggressive outlier rejection multiple times. This option should be only used with an input IRAC catalog from a different dataset that has been solved with CD.solver. Or the values of realz are very accurate.

CD.solver3 follows the algorithm of CD.solver after first estimating the time-dependent scale factor. This scale factor should be measured for data observed at HMJDs greater than 3 months away from the measured scale factors (see code for dis_scale for a list of known HMJDs).

CD.solver4 is CD.solver2 with an time-dependent scale factor estimation routine first.

Returns a matrix with columns defined as RA, DEC, angle, nums, and Residuals.

RA and DEC are both in degrees, angle in radians, and residuals in the average arcsecond residual between mean position of star and an individual detection. 'nums' is an integer showing the number of stars used to measure the other three parameters.

If CD.solver3 or CD.solver4 is used a list is returned with the first component being the previously described matrix. The second component is a vector with the values of the estimated scale factor for [3.6] and [4.5]. All other functions that take the output of CD.solver are constructed to be able to take either a matrix or a list.

Taran Esplin

## Not run: 
#This takes several minutes even with multiple cores so don't casually run
data(data1)
CDtest <- CD.solver(data1)
#compare above results with 
data(CD1)
#where a realz was used in CD.solver2

## End(Not run)