The input error_file must contain at least 4 white space delimited fields. The first is the hour angle , the second is the declination, the third is the difference between the observed and predicted position in hour angle (not angular), and the fourth is the difference between the observed predicted declination. If the system option is used, the values of the telescope pointing corrections in hour angle and declination must be in fields 5 and 6. All quantities are in decimal degrees. Lines beginning with `#' are treated as comments and are skipped. The pointerr procedure will print out a file of this format.
flex calculates the rms values of the ha and dec pointing, then fits the functional forms, and prints out the results. A file containing the residuals is printed a file with the same name as the error_file but with the extension .flx.
Certain parameters of the fit can be constrained by entering their (fixed) values on the command line.
flex tries to minimize the total angular ha+dec pointing error. If you wish to ignore the hour and or declination terms you many use the noha and nodec options respectively.
If the telescope already has pointing corrections installed and you wish to update them you can specify the system option which will add the telescope pointing coefficients used by the telescope (and found in columns 5 and 6) before computing the revised values. The output .flx file will only contain the deviations of the observed-predicted values.
You may simulate the effect of pointing errors by entering their fixed values as above (e.g. a1=23.0 a2=-.020 etc.) and adding the keyword phony. The output file is written into the usual .flx file.
The data may be plotted as a vector field (a bunch of arrows showing the magnitude and direction of the errors) if the .flx file is converted using vfield program and plotted with xplot.