User's Manual for FGS Data Reduction
(FGS Instrument Report No.23)
Patricia C. Vener-Saavedra
April 21, 1993
S.T. Holfeltz
Chapter 1
Introduction
Welcome to the rst User's Manual for FGS Data Reduction. The manual
begins with a chapter on unpacking the data; that is, transforming GEIS data
les into a viable form for the scientic reduction of the observations. The
chapters which follow will concentrate on those data reduction programs.
The impetus for this software arose with the decision to utilize the FGSs
for astrometry as well as for guidance. It was quickly realized that separate
reduction processes were necessary for scientic and engineering needs. The
software development has paralleled the STScI FGS astrometry team's work
in calibrating FGS single star transfer functions, correcting for spacecraft
jitter, and so on. The original code, consisting of several les written by B.
Bucciarelli and M. G. Lattanzi, were later integrated with new code written
mainly by S. T. Holfeltz with some input by P. C. Vener-Saavedra.
There are two types of FGS measurements which are of interest to researchers in astrometry. The rst of these is TRANSfer mode data. This
is the graphical representation of a \Transfer Function" or S-curve (i.e., the
visibility fringe produced by the Koesters prism). How an empirical S-curve
varies from a theoretical one may be an indication of some of the physical
properties of the observed object. The second type of FGS data is POSitional
mode data. This data is a series of measurements of the target object's position with respect to pre-selected reference stars.
After unpacking, the type of observation will determine which of the data
reduction programs are appropriate to be used.
1
Chapter 2
UNPACK
2.1
The Purpose of
unpack
After user-specied les have been obtained, the utility unpack can be
used to unpack and extract from those les the necessary information. The
les must be GEIS header and data les whose names will be supplied by
the user using the naming convention described below. The reason for this
particular naming convention is that this software was designed to work in
concert with the data retrieval system (Starcat) at STScI. Files thus obtained
obey the naming convention by default. (The user must also specify the data
type required to be extracted via corresponding ag sequences; see the text
below). Output lenames (described below) will be hard-coded but the user
can re-name them later if so desired.
Input lenames will be of the form:
FpppssooM.AnD - GEIS data le,
FpppssooM.AnH - GEIS header le,
where ppp represents the observation program identity, ss represents the
observation set identity, oo represents the observation identity, and n represents the FGS radial bay number. An example would be F0THX101M.A3H
which is a header le for observation program 0TH, observation set X1, observation identity 01, using the FGS in radial bay # 3 (i.e., the astrometer
FGS).
Output les include:
FpppssooM.FLAGS which contains the FGE (i.e., Fine Guidance
Electronics) status ag history,
FpppssooM.mode which contains only that data which satises the
conditions and mode specied by the user (e.g., F0THX102M.TRANS).
2
2.2
mode can be any one of POS, TRANS, WALK, SCAN, MAP, or
GSn which is the Guide Star data in POS mode,
FpppssooM.SCURn is used only for TRANS data. This (these)
le(s) contain(s) individual TRANSfer scans.
Details
The program will rst ask the user for the input GEIS data le, i.e.,
FpppssooM.AnD. Should the le reside in a directory other than where
the user is presently working, then the entire pathname must be supplied.
Consider [.mydirectory]F0THX101M.A3D. The [.mydirectory] is the
name of the directory where the le F0THX101M.A3D resides. Should the
user inadvertently type in a non-existent lename, or an incorrect pathname,
the program will prompt the user to try again.
Next the program will request the header le which must be entered in a
similar fashion (depending upon its location).
unpack now converts the GEIS les to ASCII les to be internally used.
This step does not require any further input from the user. The program
then determines the FGE values of the status ags as a function of \pixel"
number. The user will be oered the option of perusing this ag history.
In any case, the user has the option to specify the ag sequence that will
determine which measurements are kept or may request one of the following:
valid POS mode data, valid TRANS mode data, or OTHER. For any choice
other than TRANS the user will be asked if the data is Guide Star data.
(The only eect this has is in the extension of the output lename, e.g.,
FpppoossM.GS1 in the case that the FGS #1 le was Guide Star data.)
At this point, depending on which ag sequence the user has chosen, the
program will move on to the appropriate branch of commands. For the choice
of POS mode data the program creates the necessary FpppssooM.POS
data le to be used with the POS mode pipeline (Lattanzi et al. 1992a). For
the choice of TRANS mode the program goes on to determine the number of
scans and then splits the present le into separate les, FpppssooM.SCURn,
one for each of the scans determined to exist. These les, in turn, are the
input for the Transfer Function Data Reduction Package including the programs scurve and binary (Lattanzi et al. 1992b). To see examples of this
utility being run, refer to gures 2.1, 2.2, and 2.3. The rst two demonstrate unpack being used to unpack TRANSfer mode data while the last
demonstrates the same for POSition mode data.
3
Figure 2.1: Sample Execution of unpack on TRANS Mode Data
4
Figure 2.2: unpack Showing Display of Status Flag History
5
Figure 2.3: Sample Execution of unpack on POS Mode Data
6
2.3
Pertinent Subroutines
The rst of these is geis to ascii which reads GEIS formatted data les
and header les and determines (and at the option of the user will display)
values of the FGE status ag history. It then converts to ASCII format that
output data which satises user-specied criteria. Within this subroutine
are calls to other subroutines that will, if necessary, clean the observstions
somewhat.
The second most important subroutine, split, is utilized only for TRANS
mode data. It determines the number of scans and writes the measurements
for each to individual les.
The rest of this manual will deal with programs utilizing the products
(output) of unpack.
7
Chapter 3
SCURVE
3.1
The Purpose of
scurve
The program scurve uses TRANSfer scan data output from the utility
unpack, specically FpppssooM.SCURn, to create S-curve plots for the
x and y axes individually. This program can create raw S-curves, cut o
superuous ends, t a continuous piece-wise polynomial to the observations,
or create plots (see the text below for more details and options). The user
has the ability to choose where graphics output is written and if it is to be
in PostScript form.
The rst subroutine called simply informs the user what the output les
from the program scurve are called and what they contain. However this
information is written not to the screen but to a le called scurve.out that
the user can access (and re-name) later. The user will then be prompted to
name an input le and then choose an axis, x or y, for which to plot the
transfer function. Next the program will ask whether this le contains observations which have been previously processed (or merged). The answer,
yes or no, determines the point at which the next subroutine will be entered.
This subroutine converts the star selector encoder angles to x and y focal
plane co-ordinates and then reads them and their corresponding photomultiplier tube (PMT) counts into an internal le. The user is asked to declare a
lename for a corresponding plotting le. This includes the specication of a
device. (A list of options is given in the Appendix.) For example raw.plt/ps
names a plotting le produced in PostScript form which can later be sent
to any PostScript laser printer. As another example, the le /null can be
specied. This sends the plot output to the null device.
>From the raw measurement le compiled by unpack, scurve cuts o
the ends of the transfer function leaving only the innermost 0.8 arcsec of
the curve. Again the user is given the option of creating a corresponding
8
plotting le and device. The next step has the program call a subroutine
that ts a continuous piece-wise polynomial to the data. The default is a
continuous cubic polynomial with the possibility of continuous rst and second derivatives. Should the subroutine fail to encounter a sucient number
of data points to accomplish this, the user is asked to supply a lower order
polynomial. Upon successful completion, the user is once more prompted
for a plot lename and output device. The user is then given a chance to
try a dierent polynomial. The default response is no, but should the user
respond in the armative the program will again prompt for the order of the
tting polynomial and continuity as described above. Finally, a subroutine
centers the resultant transfer function at x (or y) = 0 and the nal request
for a plot lename and device are prompted. Figure 3.1 shows one possible
set of choices made while running scurve. Examples of raw and smoothed
plots produced by scurve are shown in gures 3.2 and 3.3 respectively.
3.2
Pertinent Subroutines
scurve depends on a number of subroutines to formulate the Transfer
Function and plot the result. The functions of some of these have been
briey described in the last section, and we include a short listing of the
more signicant subroutines here.
prepare scurve converts star selector encoder angles into x and y
co-ordinates and writes these and their associated PMT counts to a le
to be used internally for subsequent reduction,
make scurve creates and writes a pointwise transfer function for the
chosen axis,
cut scurve removes outer wings of the transfer function in order to
view the central portion of the curve in greater detail,
scurve interp formulates a piecewise continuous polynomial t to
the observations depending on user-specied order of the polynomial
and continuity conditions,
finish scurve centers the transfer function more precisely on the origin,
plot scurve uses pgplot to write a plot of the transfer function to
a specied device or le.
9
Figure 3.1: Sample Execution of scurve
10
Figure 3.2: \Raw" Transfer Function Plot
11
Figure 3.3: \Smoothed" Transfer Function Plot
12
Chapter 4
BINARY
4.1
The Purpose of
binary
The program binary calculates various parameters of binary stars from
FpppssooM.SCURn. These parameters include the separations on the x
and y axes, x and y, the magnitude dierence between the primary and
secondary, m, and the position angle, PA. This is accomplished by comparing the observed binary transfer function to a synthetic binary transfer
function created from a linear combination of single star reference transfer
functions (see Lattanzi et al. 1992c).
User-supplied les include the polynomial les as output by the program
scurve for both the observed and single star reference transfer functions.
The latter are obtained from the STScI calibration database (CDB) while
the former are derived from the program scurve. A GEIS header le for the
observed binary must also be supplied if the user wishes the program to nd
the position angle (this le provides the V3 roll as well as RA and Dec of the
V1 axis). In the case where the user wants position angle, the program will
prompt for axis as determination of the position angle requires resolution of
the binary in both the x and y axes.
binary rst introduces itself to the user, explaining its function and detailing what les and information must be supplied to it by the user. Next
it sets the ranges of values to be searched and the corresponding step sizes
for oset, separation, and (m =) Dmag and oers the user the option
of plotting the reference and observed transfer functions (to be plotted via
PGPLOT at an output device of the user's choice). Next polynomial representations of the above transfer functions are used to determine the best
t oset, separation, and Dmag of the observed binary. A synthetic binary
transfer function polynomial is generated from the reference single star to
nd the oset that produces the highest correlation with the observed bi13
nary. Finally, the computed values of separation and Dmag (for the x and
y axes separately) are used with the predicted roll of the telescope to nd
the position angle of the binary. Figure 4.1 shows an example of a run of
binary.
At this point it should be noted that one may nd it useful to run the
program merge in advance of binary. The purpose of which is to increase
the signal-to-noise ratio of a target by merging several scans of that target.
4.2
Pertinent Subroutines
binary uses the following four subroutines to accomplish its tasks:
intro, as may be construed from the name, provides an introduction
to the utility, its uses and its limitations,
setdefs sets up default values for ranges to be searched and step sizes
for oset, separation, and Dmag,
polyfit generates the matrix of binary templates from a polynomial
representation of the single star reference and observed transfer functions,
pointfit via a least squares approach uses the template with the highest correlation to the observed transfer function to determine the bestt oset, separation, and Dmag.
14
Figure 4.1: Sample Execution of binary
15
Chapter 5
POS
5.1
The Purpose of
pos
POS calculates the instantaneous target position in various co-ordinate
systems (e.g., V1, V2, V3 space, FGS image space, and so on). It can correct
for OFAD (optical eld angle distortion), velocity aberration, FGS to FGS
misalignments, and so forth. The data it uses derives from the utility unpack, specically the les FpppssooM.POS and FpppssooM.GSn. The
user must also supply the appropriate GEIS header le (from which is obtained predicted attitude data and predicted HST orbital data, V3 roll, R.A.
and Dec. of V1) as well as an Earth ephemerides le. This last le can be
obtained from percy, a program maintained by the Moving Object Support
System. pos will also ask the user to specify FGS number, year, and whether
or not this is a leap year. As an alternative to using predicted HST orbital and
attitude data, the user may use HST attitude data supplied by Observatory
Monitoring System (OMS) derived from the spacecraft's telemetry stream or
HST orbital data from OMS or DMF. Output is written to Fpppssoom.dat.
When the data is guide star data, it will generally be used for the correction of spacecraft jitter in the program scurve. Figure 5.1 shows a sample
execution of this program.
5.2
Pertinent Subroutines
There are three subroutines called upon by pos that are of major importance and we describe them here (see Ta 1990, Holfeltz, et al.).
fgs to veh converts star selector encoder angle to FGS image space,
FGS de-magnied image space, and (V1, V2, V3) vehicle space,
ofad applies a correction for the Optical Field Angle Distortion,
16
Figure 5.1: Sample Execution of pos
17
vel aberr applies a correction for the dierential velocity aberration
in V2, V3 space. (Note that this is not dierential in time but with
respect to V1.)
18
Appendix A
Graphics Devices
Several of the utilities discussed in this manual are capable of producing plots
via PGPLOT. When this is so the user is given the opportunity to choose
the device to which this output will be directed. These choices are listed and
described in the table below.
/GF
/FILE
/NULL
/PS
/PRINTRONIX
/RETRO
/TEK4010
/TFILE
/TK4100
/VPS
/VT125
/VT340
/X11
GraphOn Terminal
generic le
don't keep it
PostScript le
landscape
Table A.1: Graphics Device Types
19
Bibliography
[1] Ta, L.G. 1990, \The Astrometric Calibration of the Hubble Space
Telescope Fine Guidance Sensors" Ap. J. 365: 407-418.
[2] Lattanzi, M. G., Bucciarelli, B., Holfeltz, S.T., and Ta,
L.G. 1992a, \Astrometry with the HST Fine Guidance Sensors"
Developments in Astrometry and Their Impacts on Astrophysics and
Geodynamics. IAU Symposium No. 156 Shanghai, China.
[3] Lattanzi, M.G., Bucciarelli, B., Holfeltz, S.T., and Ta, L.G. 1992b,
\The Analysis of HST Fine Guidance Sensor Transfer Functions"
Complementary Approaches to Double And Multiple Star Research.
IAU Symposium No. 135 Atlanta, Georgia, USA.
[4] Lattanzi, M.G., Bucciarelli, B., Holfeltz, S.T., and Ta, L.G. 1992c,
\Fine Guidance Sensor Instrument Report: The STScI Transfer Function Mode Data Reduction Package" Fine Guidance Sensor Instrument
Report #19.
[5] Holfeltz, S.T., Lattanzi, M.G., Bucciarelli, B. and Ta, L.G. 1993, \The
STScI POSitional Mode Data Reduction Code" Fine Guidance Sensor
Report #24.
20