iraf.net

From: Andrea Dupree <adupree@cfa.harvard.edu>
Date: July 6, 2005 9:19:12 PM GMT+02:00
To: fpierfed@noao.edu
Cc: dupree@cfa.harvard.edu
Subject: splot questions
HI Francesco,It's me again with splot questions...1. [How] can I change the format in the splot.log output to
carry 8 figures instead of 7?2. Is there a straightforward way to automatically (not interactively)
apply an offset to a spectrum (in splot?).. I know there is a offset
using cursor u, then z but that involves an interactive selection.3. Inserting comments in to a splot.log file using
:# comment to be addeddoes not seem to work in a gterm window... should it? should
the comment be in brackets? should it be prefaced with :/# ???
thanks, and belated happy fourth of July...
Andrea
--
**********************************************
Andrea Dupree
Smithsonian Astrophysical Observatory
Harvard-Smithsonian Center for Astrophysics
60 Garden Street MS-15
Cambridge, MA 02138 USA
email: dupree@cfa.harvard.edu
Tel. 617.495.7489 FAX 617.495.7049
**********************************************
From: Frank Valdes <valdes@noao.edu>
Date: July 8, 2005 9:45:32 AM MST
To: adupree@cfa.harvard.edu
Subject: Re: splot questions
Hi Andrea,Francesco is no longer at NOAO. He forwarded your message to
iraf@noao.edu. Below are my answers to your questions.Yours,
Frank>
> HI Francesco,
>
> It's me again with splot questions...
>
> 1. [How] can I change the format in the splot.log output to
> carry 8 figures instead of 7?
>
>I'm afraid this is hardwired in the program. Which particular
measurement result are you concerned about? Usually this problem
would occur with very high resolution data. There are tricks you can
apply to fool with the wavelength scale. One easy one is to use '$'
to convert the scale to pixels and measure things in pixels. Then
you would use the WCS information in the header to convert to
wavelength. For a linear WCS (CRVAL/CRPIX/CD1_1 type of keywords it
is easy to calculate directly. Or you can use WCSCTRAN to convert
pixels ("logical" coordinate system) to wavelength ("physical"
coordinate system).Another approach is to fudge the zero point so that in the range of
data that you want to
measure the numbers have fewer high order digits. You can use
SPECSHIFT (see below).
For example if you data is between 4200 and 4400 you could shift by
-4000 and then
your measured feature might be 222.1234 instead of 4222.123. Or if
you shift by -4200 you would get 22.12345.
> 2. Is there a straightforward way to automatically (not interactively)
> apply an offset to a spectrum (in splot?).. I know there is a offset
> using cursor u, then z but that involves an interactive selection.
>
>To automate SPLOT you should use the BPLOT command. You make a file
that has the cursor coordinates and keystrokes you want. Let me know
if you need an example but check the help page for BPLOT.However, if you just want to apply a zeropoint shift to the
wavelengths you can use SPECSHIFT. Or you can just use HEDIT to
change the appropriate CRVAL1 value (assuming the dispersion is a
simple linear one).
> 3. Inserting comments in to a splot.log file using
> :# comment to be added
>
> does not seem to work in a gterm window... should it? should
> the comment be in brackets? should it be prefaced with :/# ???
> thanks, and belated happy fourth of July...
>
>In my test, if I understand your question, it works fine. When I
have the graph up and a cursor I just type ":# This is a
test" (without the quotes). Note that this can be put in the cursor
file used with BPLOT to annotate the log. Since this independent of
position of the cursor you just need
to add lines to the cursor file such as:# This is a test.
From: Andrea Dupree <adupree@cfa.harvard.edu>
Date: July 8, 2005 11:53:42 AM MST
To: Frank Valdes <valdes@noao.edu>
Subject: Re: splot questions
HI Frank...
Thanks!!! for your answers... actually I had gone into hedit and
did the offset but had not come across specshift....(Problem is that I am working at 10830Angstroms, and by the time
splot deals with that it has used up 5 of its 7 'spaces'... for some
reason also it seems to think that the wavelength is 'meters"... don;t
know how to change that... )And for some reason, inserting a comment appears to work now... don;t
know why!...thanks again,Andrea--
**********************************************
Andrea Dupree
Smithsonian Astrophysical Observatory
Harvard-Smithsonian Center for Astrophysics
60 Garden Street MS-15
Cambridge, MA 02138 USA
email: dupree@cfa.harvard.edu
Tel. 617.495.7489 FAX 617.495.7049
**********************************************
From: Frank Valdes <valdes@noao.edu>
Date: July 8, 2005 12:44:58 PM MST
To: Andrea Dupree <adupree@cfa.harvard.edu>
Subject: Re: splot questions
If you send me a header dump I can tell you why it thinks the units
are meters. It is an easy HEDIT to fix it so you see Angstroms or
nm, etc. I guess you realize that simply changing units doesn't
solve the significant digits problem unfortunately. SPLOT, and other
spectral tasks, really need to be changed in some clever way to deal
with output format issues.Frank
From: Andrea Dupree <adupree@cfa.harvard.edu>
Date: July 8, 2005 1:14:25 PM MST
To: Frank Valdes <valdes@noao.edu>
Subject: Re: splot questions
HI..Here is the header... I can't see meters in it?...
The data arrived in a circuitous way... from the NOT telescope
in the canaries...thanks, andrea
--
**********************************************
Andrea Dupree
Smithsonian Astrophysical Observatory
Harvard-Smithsonian Center for Astrophysics
60 Garden Street MS-15
Cambridge, MA 02138 USA
email: dupree@cfa.harvard.edu
Tel. 617.495.7489 FAX 617.495.7049
**********************************************
SIMPLE = T / Standard FITS
format BITPIX = -32 /
Four float bytes per pixel: LSB is las NAXIS =
3 / Three-dimensional image NAXIS1
= 1114 / Number of pixels along
axis NAXIS2 = 1 / Number of
pixels along axis NAXIS3 = 2 /
Number of pixels along
axis
CRVAL1 = 10796.538874 /
Initial value of the axis CRVAL2 =
1.000000 / Initial value of the axis CRVAL3
= 0.000000 / Initial value of the
axis CDELT1 = 6.654020e-02 / Step size
along axis CDELT2 = 1.000000e+00 /
Step size along axis CDELT3 = 1.000000e
+00 / Step size along axis CRPIX1
= 1.000000 / Reference pixel
number CRPIX2 = 1.000000 /
Reference pixel number CRPIX3 =
1.000000 / Reference pixel number CTYPE1 =
'Wavelength' / Name of axis
CTYPE3 = 'Variance' / Name of
axis BUNIT = 'ADU' / Physical
unit of pixel
values
DETECTOR= 'ASTROMED-3200 camera with EEV3
P88300 1242x1152 CCD' CCDCOLS =
1152 / Number of CCD columns CCDROWS =
1242 / Number of CCD rows COMMENT
VOD=16.48 VRD=4.00 VOG=-10.05 VSSL=-6.50
VSSH=-2.01 COMMENT VRSPH=-4.97 VROH=-2.01 VIMH=-3.03
VSTOH=-3.03 COMMENT VRSPL=-12.26 VROL=-12.33
VIML=-12.35 COMMENT DIDS=-2.0
DWELL=21 RSTCCD=5 CCDTEMP
= 182.362896 / CCD temperature in K at read
out CCDTEMP1= 181.410216 / CCD temperature
sensor 1 CCDTEMP2= 183.315575 / CCD
temperature sensor 2 DEWATEMP=
89.652147 / CCD dewar temperature AMBTEMP
= 285.803823 / Ambient temperature in K at read
out GAIN = 1.930000 / CCD gain factor in e/
ADU RON = 2.600000 / CCD
readout noise in ADU DATE =
'05/03/2001' / Local date the file was
written
OBSERVAT= 'LaPalma' /
Observatory name TELESCOP= 'Nordic Optical
Telescope' / Data acquisition telescope LATITUDE=
'+28:45:20.50' / Latitude of telescope in degrees
LONGITUD= '-01:11:32.00' / Longitude (east>0) in
hours ALTITUDE= 2382.000000 / Altitude of
telescope in
meters
OBSID = 'Dupree' / Andrea Dupree,
CfA OBSERVER= 'Dupree Ilyin' /
Observer(s) name/identification ORIGIN =
'4A' / Data acquisition software
INSTRUME= 'SOFIN spectrograph' / Data acquisition
instrument REGION = 'HeI 10830' /
Spectral region name GRATPOS =
8230.000000 / Dispersion unit position PRISPOS
= 1650.000000 / Cross-dispersion unit
position FILTER = 3 / 3500 -
6000 Blue CAMERA =
2 / Optical camera number SLIT
= 77.000000 / Entrance slit width in
microns
IMAGETYP= 'object' /
Object exposure OBJECT = 'HD
47667' / Object observed
BARVEL = -9864.700283 / Barycentric velocity m/
s DUIRVEL = 136.740414 / Duirnal
velocity m/
s
COUNTS = 44.878000 /
Photon counter/1000 DATAMIN = 0.000000e
+00 / Data minimum in real scale DATAMAX =
0.000000e+00 / Data maximum in real scale
HISTORY Column BSS27162.GSS subtracted from
CCD27162.FIT HOST =
'EXT27162.SUB' / The original image name
HISTORY Ratio of 2 files ext27162.col
ext27321.nor HISTORY Difference of 2
files ext27162.dvd ext27162.bgn HISTORY
Weighted extraction: alpha=1e+10 median=15
ksi=3 HISTORY Ratio of 2 files ext27162.opt
ext27168.nor ORDER =
21 / Spectral order number FWHM
= 4.308270 / FWHM of the
order HISTORY Wavelength calibration image is
ext27161.std IMAGENUM= 27162 /
Image number HISTORY Radial velocity
EARTHVEL=-9727.96 is removed and error is not added DATE-OBS=
'09/12/2000' / UT of midexposure date
UT-OBS = '04:05:18.4' / UT of midexposure
time HISTORY Sum of 5 files 21x27162.hwl
21x27164.hwl 21x27166.hwl 21x27169.hwl HISTORY
21x27172.hwl H
ISTORY 21X27162.SUM smoothed by weighted polynomial of 3
degree HISTORY Sigma clipping above=0
below=2 HISTORY Curve is in:
21X27162.POL ENDFrom: Frank Valdes <valdes@noao.edu>
Date: July 8, 2005 1:57:20 PM MST
To: Andrea Dupree <adupree@cfa.harvard.edu>
Subject: Re: splot questions
Hi Andrea,The old FITS standard said that if CTYPE1 was wavelength then the
units were to be assumed to be meters. In virtually all cases this
has caught people, including the originators of your data since
obviously the units are supposed to be Angstroms. In the latest
version SPLOT would assume Angstroms. I don't have immediate access
to a version that you are using but I think if you delete CTYPE1 it
will assume Angstroms. You can also addhedit <spectrum> cunit1 Angstromsand it should show you Angstroms.
Frank