Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/sas/6.5.0/doc/lcplot.ps.gz Дата изменения: Tue Aug 16 02:31:08 2005 Дата индексирования: Tue Oct 2 01:53:57 2012 Кодировка: Поисковые слова: http astrokuban.info astrokuban

XMM-Newton Science Analysis System Page: 1
lcplot
August 15, 2005
Abstract
lcplot produces plots of the net source and background time series for a single time
series FITS le or for several time series FITS les produced by elcbuild for EPIC
and omlcbuild for the OM. Furthermore it performs Chi Square and Kolmogorov-
Smirnov tests on the light curve to determine whether the source is variable.
1 Instruments/Modes
Instrument Mode
EPIC MOS IMAGING, TIMING
EPIC PN IMAGING, TIMING
OM FAST
2 Use
pipeline processing yes
interactive analysis yes
3 Description
lcplot produces plots of time series measured either by EPIC or by the OM. These plots show both net
and background count rates. lcplot can be used to display a single time series directly or the multiple
time series, produced by elcbuild for EPIC and by omlcbuild for the OM, and supplied as FITS les.
Alternatively the task omthlcplot can be used to produce rate plots of the OM tracking star's intensities.
The input FITS les must contain net count rates, background count rates and the associated errors.
The time series must follow a regular binning scheme (i.e. equispaced time bins) [1]. The bin width is
given by the keyword TIMEDEL.
Each light curve (net and background) of the input le is plotted on a separate page. For instance, in a
PPS use, EPIC time series are given in three energy bands, so the plots will be composed of three light
curve pages. On the other hand, an OM FITS le which contains only one light curve, will give a single
page plot.
xmmsas 20050815 1803-6.5.0

XMM-Newton Science Analysis System Page: 2
The task performs 1) the Chi Square and 2) the Kolmogorov-Smirnov tests on the total counts. lcplot
tests the null hypothesis (the source is not variable) using 1). A Chi Square t is carried out on the count
distribution and the constant distribution whose value is equal to the observed mean (both statistics and
the probability are computed) using 2). A Kolmogorov-Smirnov test is carried out on the cumulative
probability functions of the observed count distribution and the theoretical Gaussian distribution whose
mean and variance are both equal to the observed mean count.
Results are written on a single page after the light curve(s).
Plots give time after TSTART (keyword in the FITS le) on the X-axis against count rates on the Y-axis.
Please note, bins with a low exposure and therefore large errors are not plotted.
lcplot allows the user to input the device type (e.g. Xwindow, Postscript, ...) for output. The y-scale
of the background counts can be forced to be the same as the source counts by setting the parameter
`bkgdyscale' to `yes'. Otherwise the y-axis is scaled to the most appropriate value for the background
counts.
As described in [1], light curves should be rebinned if necessary in order to limit the number of plotted
data points to 500.
3.1 Examples
To create a postscript plot, called plot.ps, with both net and background count rates and the results of
the Chi Square and Kolmogorov-Smirnov tests, from the FITS le 'lightcurve.ds', where the counts are
binned up by a factor 10, a command such as,
lcplot set="lightcurve.ds" plotdevice="/VPS" plotfile="plot.ps" binsize=10
can be used.
4 Parameters
This section documents the parameters recognized by this task (if any).
Parameter Mand Type Default Constraints
set yes dataset No default
Name of the FITS le containing the time series. Case sensitive.
plotdevice no String \/PS" Constraint : valid de-
vice type
Not case sensitive.
plotle no String \plot.ps\ less than 1024 char.
Name of the output le containing the plot. Case sensitive.
binsize no Integer 5 Constraint : positive
value
bkgdyscale no boolean no yes/no
xmmsas 20050815 1803-6.5.0

XMM-Newton Science Analysis System Page: 3
To scale the y-axis of the plot of the background counts to be the same as the y-axis of the plot of the
source counts
5 Errors
This section documents warnings and errors generated by this task (if any). Note that warnings and
errors can also be generated in the SAS infrastructure libraries, in which case they would not be docu-
mented here. Refer to the index of all errors and warnings available in the HTML version of the SAS
documentation.
Incorrect RATE extension (error)
Incorrect RATE extensions (see Developer's notes for more details)
Keyword TIMEDEL missing (error)
The keyword TIMEDEL is missing
Existing output le cannot be overwritten (error)
The output le exists, and cannot be overwitten because the SAS CLOBBER variable is set
to 0
Unavailable default device type and/or output le (error)
Unavailable default device type and/or output le
Invalid input le (error)
Invalid input column dimensions (the four columns RATE, ERROR, BACKV, BACKE contained in
the FITS le must be arrays of dimension Nlightcurve  Nbins)
Unavailable device type and/or output le (warning)
corrective action: The device type and output le are set to the defaults.
Number of data points to plot is higher than 500 (warning)
corrective action: Increase binsize value until the number of points to plot is lower than 500
[1]
Too small number of bins (warning)
Number of valid bins is lower than 2. Cannot perform tests
corrective action: No variability tests are delivered
Mandatory keywords are missing from the input le (warning)
Some mandatory header keywords are missing or inconsistent
corrective action: A warning is raised and the task tries to proceed
6 Input Files
1. EPIC or OM FITS Time Series [1] (generated by omlcbuild or elcbuild).
xmmsas 20050815 1803-6.5.0

XMM-Newton Science Analysis System Page: 4
7 Output Files
1. EPIC or OM Postscript Source Time Series [1] (Deliverable PPS product)
8 Algorithm
Subroutine lcplot
Recover parameters (File names, Device type, Binning factor, Ouput file)
Read source time series FITS file :
Get dataset and table.
Check important keyword consistency (notably TSTART and TIMEDEL).
Call a warning or error if necessary.
Recover all light curves included in table (in PPS use, 3 for EPIC, 1 for OM) :
Net source rates and errors & background rates and errors are recorded in arrays
of dimension Nlightcurve * Nbins.
Open plot : create view surface in specified device format (check its validity).
For each light curve :
Delete gaps in data (when the IEEE NaN constant is found).
Increase binsize if necessary to limit the number of points to 500.
Rebin rate and error columns by binsize
Create a labelled plot page separated into two panels in which time is graduated
on X-axis (see Comments), and count rates on Y-axis (in counts/sec).
Draw net source and background time series on different panels.
Reconstruct net source counts and total counts rebinned :
Perform variability tests on rebinned net source counts :
Calculate mean count rate and variance.
Test the null hypothesis (the source is not variable) by :
- Processing a Chi Square fit between the count distribution and a
constant distribution whose value is equal to the observed mean (statistic
and probability are computed).
- Processing a Kolmogorov-Smirnov test between the cumulative probability
functions of the observed count distribution and a theorical Gaussian
distribution whose mean and variance are both equal to the observed mean
count.
End for
Write variability test results on a single plot page.
Release memory.
xmmsas\_20050815\_1803-6.5.0

XMM-Newton Science Analysis System Page: 5
Close plot.
End subroutine lcplot
9 Comments
 lcplot only produces plots of light curves measured by EPIC or OM. We note that no RGS
Source Time Series FITS le is specied in [1].
 Error rates are rebinned following a quadratic scheme (Central limit theorem).
 Time intervals all have the same length, the rst one is [TSTART , TSTART + Binsize*TIMEDEL]
where TIMEDEL is the bin width read from the le [1].
 We note in [1] that the keyword TIMEDEL is specied in MJD. We have assumed it to be
given in seconds.
 Time on the X-axis is given as T-TSTART in seconds.
 The mean, variance, Chi Square, etc are provided with six signicant digits.
 A Chi Square statistic close to the bin number and a small Kolmogorov-Smirnov statistic
with probabilities higher than 0.01 indicate that the source is not variable.
 The output devices are those of the PGPLOT library. The device list is governed by the
drivers available to PGPLOT.
10 Future developments
 A parameter Time Unit could be added to the parameter le. It would allow a choice of
dierent time axis units (e.g. Hours after TSTART).
 The results of the statistical analysis could be saved as keywords in the header of the light
curves source FITS le. Example, for the rst energy band, one could add :
AVRGE_1 = .............. / Avg, count/s in light-curve
FREXP_1 = ............. / Avg. Frac exposure in light-curve
VAROB_1 = ............. / observed variance
VAREX_1 = ............. / expected variance
THRDM_1 = ............. / third moment
MININ_1 = ............. / Minimun Intensity
MAXIN_1 = ............. / Maximun Intensity
EXVAR_1 = ............. / excess of variance
CHI2_1 = ............. / Chi squared
CHI2PR_1= ............. / Chi squared probability
KS_1 = ............. / K-S test value
KSPR_1 = ............. / K-S probability
RMS_1 = ............. / RMS variability
AVRGE_E1= ............. / Error Avg count/s
VAROB_E1= ............. / Error observed variance
VAREX_E1= ............. / Error expected variance
CHI2_E1 = ............. / Error Chi2
RMS_E1 = ............. / Error RMS variability
xmmsas 20050815 1803-6.5.0

XMM-Newton Science Analysis System Page: 6
10.1 CAL usage
 None
References
[1] SSC. XMM Survey Science Centre to Science Operations ICD for SSC Products. Technical Report
XMM-SOC-ICD-0006-SSC Issue 2.1, SSC, Mar 2000.
xmmsas 20050815 1803-6.5.0