NAME
xrtpdcorr -- Calculate, subtract or adjust the bias for data taken in Photodiode mode
USAGE
xrtpdcorr [parameter = < value >]
DESCRIPTION
'xrtpdcorr' checks if the bias subtraction has been applied on-board to
the Swift XRT Photodiode mode data and, if not, calculates and subtracts the bias. For the
Low Rate Photodiode only, the task estimates the bias also if the bias subtraction has been already
applied on-board and, if necessary adjusts the bias subtraction.
The on-board software can be set to either subtract the bias on-board before sending down
the data (default) or to send down the data without bias subtraction.
In the latter case the bias can be subtracted on-ground in several different ways:
-
User input bias value.
The user can input a bias value through the parameter 'bias', setting the
parameter 'method' to 'CONST'.
If the parameter 'bias' is set to a negative value, the task will read the
bias value from a file specified through the parameter 'biasfile'.
If the 'biasfile' parameter is set to CALDB (default), the bias value is
read from the file stored in the Calibration Database, if it is set to a
value different from CALDB, the task will look for a file provided by the
user. This file must have the same format as the bias file stored in CALDB.
-
CALDB input bias value
(The parameter 'bias' set to a negative value, 'biasfile' set to 'CALDB', 'method'
set to 'CONST').
A constant bias value has been evaluated by on-ground
analysis of calibration data and is stored in a CALDB file that will
be periodically updated using in-flight calibration.
-
Calculate bias value using a statistical method.
A subset of telemetered PHA values is used to calculate the bias.
For the Low Rate Photodiode, the bias is calculated using the last 20 pixels
telemetered in each frame and stored in the HK Header file.
For the Piledup Photodiode, the bias is calculated using
the PHA values of all the pixels below threshold. 'xrtpdcorr' creates an
histogram of the PHA values and evaluates the bias by computing the mean value,
or fitting them to a Single Gaussian or to a Double Gaussian, depending on the
value of the parameter 'method'.
It is possible to perform a sigma clipping setting the parameter 'nclip' to a
number greater than 0 and the number of sigma to clip can be input through
the parameter 'nsigma'.
If it is not possible to perform the fit, the task uses the bias value from CALDB,
giving a warning to the user that it has been done so.
The 'nframe' input parameter allows to calculate the bias value
and apply the bias correction separately for a set of consecutive frames.
For Low Rate Photodiode only, the bias is estimated also if the bias
correction has been applied on-board. The task calculates the median value of the 20 pixels
telemetred and, if the bias value obtained differs from the one subtracted on-board more than a
quantity set by the user through the parameter 'biasdiff', the bias correction is adjusted.
After the correction is applied the keyword XRTPHACO, in the EVENTS extension header,
and the extension BIAS are added or updated. The keyword is set to 'T' to indicate
that the correction was done and the BIAS table stores the list of the bias values
subtracted for each frame. Only for Low Rate Photodiode event files with bias subtracted on-board,
the BIAS table stores for each frame the list of the bias values used to adjust the
bias correction.
The input event file can be supplied in the compressed format except
in the case of outfile=NONE when the input file will be overwritten.
PARAMETERS
- infile [file name]
- Name of the input event FITS.
- hdfile [file name]
- Name of the input Housekeeping header Packets FITS file. This file contains the 20 pixels used
to calculate the bias for the LR. (Necessary only for the
LR mode, set to 'NONE' for PU mode).
- outfile [file name]
- The output file name.The value 'NONE' will cause the input file to be overwritten.
- bias [integer]
- Bias value. If a negative value is input, the bias value is read from the file specified in
the 'biasfile' parameter.
- (nframe = 0) [integer]
- Number of consecutive frames included in the bias calculation. If set to 0
or to a negative value, all frames available are taken into account in
the bias calculation.
- (nevents = 20) [integer]
- Minimum number of Photodiode events with DN value under the Low Level Discriminator
included in the bias calculation when the chosen method is either MN or a Gaussian fit.
(see 'method' parameter).
- (nclip = 1) [integer]
- Number of iterations to compute the Photodiode bias value. The value for this paramter
is is used for the MN,SG or DG methods (see 'method' parameter).
If set to 0 no clipping will be done.
- (nsigma = 2) [integer]
- Number of sigmas used in the sigma clipping algorithm (Used only if 'nclip' > 0).
- thrfile [file name]
- Name of the input file containing the bias threshold value as a function of
the CCD substrate voltage. If the parameter is set to CALDB, the file is read from the calibration
database. If the parameter is set to NONE, the bias threshold value must be input
via the parameter 'biasth'.
- (biasth = 300) [integer]
- The event threshold used to select the events to compute Photodiode bias
value with MN,SG or DG methods (see 'method' parameter).
This parameter is used only if 'thrfile' is set to NONE.
- (biasdiff = 20) [integer]
- Threshold value for bias adjustment (only for Low Rate Photodiode mode).
- (biasfile = CALDB) [file name]
- Filename containing the bias value for the Photodiode mode.
If the parameter is set to CALDB (default), the bias value is
read from the file stored in the Calibration Database, if it is set to a
value different from CALDB, the task will look for a file provided by the
user. This file must have the same format as the bias file stored in CALDB.
- method [string]
-
Methods to calculate Photodiode Bias value. The methods allowed are: MN,SG,DG or CONST.
'MN' calculates a mean from the histogram generated taking number of events,input through
'nevent' parameter, under the threshold. 'SG' or 'DG' fit the histogram with a single
gaussian or a double gaussian. 'CONST' uses a constant bias value taken either from
the PD Bias Calibration file or from the value input in the parameter 'bias'.
For the 'MN' ,'SG', 'DG' methods a sigma clipping can be applied
setting the 'nclip' parameter value > 0.
- (fittol = 1.E-8) [real]
- Relative tolerance of fit error computing PD bias value with SG or DG method
(see 'method' parameter).
- (clobber=no) [boolean]
- If set to yes, overwrite the output file.
- (history=yes) [boolean]
- If set to yes, write parameter values and other information in HISTORY blocks.
- (chatter = 2) [integer]
- Chatter Level (min=0, max=5).
EXAMPLES
- Apply bias correction on the file named sw00000530000xlrb0po_uf.evt
using a bias value of 130 and writing the output in the already existing
file 'outfile.fits' that will be overwritten.
> xrtpdcorr infile=sw00000530000xlrb0po_uf.evt hdfile=sw00000530000xhd.hk
bias=130 method=CONST outfile=outfile.fits clobber=yes
- Compute the bias value by fitting the histogram of the PHA values of the
last 20 pixels of each frame, included in the housekeeping file
sw00000530000xhd.hk, with a Single Gaussian; apply the bias correction on
the event file sw00000530000xlrb0po_uf.evt and write the output in the
file named outfile.fits.
> xrtpdcorr infile=sw00000530000xlrb0po_uf.evt thrfile=CALDB
hdfile=sw00000530000xhd.hk method=SG outfile=outfile.fits
BUGS
SEE ALSO
LAST MODIFIED
July 2007