Ian M George: RXTE Software & Analysis Crib Sheets
These pages are unofficial in all respects.
They are primarily simply notes to myself regarding the installation
on .
The information is not guaranteed to be correct or up-to-date.
Rate-Selected Obsids
After
running rex
to make light curves, one can use these light-curves to
make a new obsid.lis file containing
only those observations for which the mean
count rate lies which a given range.
This is achived using the
find_lc_rex_ratecut.csh script.
- As an example, lets assume you want a list of obsids for which the
the mean count rate
in a previously constructed set of light curves
std_layer1_ch8-27_net.lc
lies in the range 13.5 to 14.5 counts per second.
- Assuming you ran rex from the
{wherever}/ngc999/work/ directory
(as suggested),
then the files required
(ie the std_layer1_ch8-27_net.lcs)
will be in directories
{wherever}/ngc999/work/12345-xx-yy-zz/.
In the directory {wherever}/work/
- Make sure the obsid.lis is correct.
For instance, using the examle from
earlier
cp obsid_P12345_epoch4.lis obsid.lis
- Then simply
./find_lc_rex_ratecut.csh std_layer1_ch8-27_net 14.0 1.0
- The value 14.0 on the command line
refers to the centre of the range of count rates wanted.
- The value 1.0 on the command line
refers to the full width of
the range of count rates wanted.
Assuming at least dataset is found within the requested range,
and output file called rex_ratecut.out
will be written.
- Obviously you probably want to rename this to something descriptive
for future use
(& to stop it being over written)
mv rex_ratecut.out std_layer1_ch8-27_net_ratecut_14.0_1.0_obsid.lis
The next step is probably to use this new list of obsids to
construct a mean Spectrum for this Rate-level.
More Sophisticated Rate-Selected Datasets
After
running rex
to make light curves, one can use this light-curve as a basis
for further rate-selected analysis.
This is achived in several steps
- First make a new gti file based on the
desired rate criteria using the ftool
maketime
- For example, say one wants to make the new gti
for times when the count rate in the file
std_layer1_ch8-27_net_16.lc_snip
was in the range
7 to 8 counts/s/PCU:
-
maketime infile=std_layer1_ch8-27_net_16.lc_snip outfile=rate_0708.gti expr="(RATE1.GE.7).AND.(RATE1.LT.8)" name=NAME value=VALUE time=TIME compact=no clobber=yes
This is what we want, however I'm afraid there is a bit
of messing around we have to do associated with zeropoint
of the above gti and that which we'll
need later....
- Find the required values & fix the zeropoint in our new
gti using a few ftools
- Note some of the header keywords in our gti
file (1st extension)
-
fdump rate_0708.gti+1
Should see something like this:
TIMEUNIT= 'd ' / Units for header timing keywords
TIMEZERI= 10405 / Zero-point offset for TIME column
TIMEZERF= 0.6860134540711442 / Zero-point offset for TIME column
COMMENT TIMESYS keyword is not currently set. If the input lightcurve
COMMENT contains the header keyword MJDREF, the time in the xronos
COMMENT output is in TJD (JD-2440000.5)
- We're going to be combining our gti file
with another ("standard") gti (made by
hackedrex_v2 below).
The latter files will have a zeropoint given by
TSTART = 9.095198900000000E+07 / As in the "Time" column
TSTOP = 9.098014900000000E+07 / As in the "Time" column
MJDREFI = 49353 / 1994.0(UTC) expressed in TT (integer part)
MJDREFF = 6.965740740000E-04 / 1994.0(UTC) expressed in TT (fractional part)
COMMENT MJD = JD - 2400000.5
TIMESYS = 'TT ' / XTE time will be TT (Terrestrial Time)
TIMEZERO= 3.37843167E+00 / Clock correction
TIMEUNIT= 's ' /
- So we need to calculate and fix the zeropoint in
our gti file to be compatible with the
latter.
- In the above example, our gti is
specified in seconds from the Julian day
2440000.5 + 10405.6860134540711442
- The "standard" gti will be specified in seconds
from the Julian Day
2400000.5 + 49353.0006965740740000
- ie Our gti has a zeropoint
which is LATER by
1052.6856199 days = 9.0952036e7 s
- Thus we can use the ftool fmodhead
to add an appropriate offset to our gti
- Now run hackedrex_v2. This is basically the
same as the regular ftool rex, except that
one is prompted for gti, which (internally)
is merged (using the ftool mgtime) with the
"standard" gti that rex
makes for each obsid (called "basic.gti"). This
merged gti is of course then used for the
extraction of the spectral data.
Example:
./hackedrex_v2 -d ../P40151 -l 1 -c 7-23 on these
Target [all]:
Printmode, l (ligh curve), s (spectra) [both]:s
Our rate-selected GTI file: rate_0708.gti
Root name for products []: rate0708
which will (should) produce source & background spectral files
rateo708_layer1_*.pha
RXTE Analysis Crib sheets
Other CribSheets
Ian M George Home Page