| Parameter | Mand | Type | Default | Constraints |
| In general for this task only one filename need be provided and the rest are inferred by default using the provided filename as a template. The rules governing this behavior are intended to follow what the user would instinctively expect and as a result are somewhat complicated. The descriptions below focus on the meaning of each filename parameter, leaving the exact details of the inferred default value to be explained in the algorithm section. | ||||
| evlist | no | dataset | ||
|
The filtered event list. It must contain the EVENTS table
and exposure map extensions. Not modified. The default value
infers the EVENLI file.
| ||||
| srclist | no | dataset | ||
|
The source description list. It must contain the SOURCES
table and the relevant selection region extensions. Not modified.
The default value infers the SRCLI_ file.
| ||||
| rmfset | no | file | ||
|
The response matrix file (OGIP-compliant format). Either created or
modified depending on parameter newrmf. When modifying an
existing response matrix (newrmf=no) an explicit value for
this filename must be provided, which then becomes the template from
which other filenames may be inferred. For a new response matrix
the default value infers the RSPMAT file.
| ||||
| witharffile | no | boolean | no | |
|
Switch to enable the creation of two independent response files ARF and RMF.
| ||||
| arfset | no | file | ||
|
The ARF file name (OGUP-compliant format).
| ||||
| emax | no | real | 2.8 | non-negative |
|
When creating a new response matrix (newrmf=yes), this is
the upper edge of the highest incident-energy bin to be included.
When overwriting part of an existing matrix (newrmf=no),
this identifies the upper incident-energy bin in the interval to be
recomputed. This value will be raised automatically as necessary to
maintain the contiguity of the bins, and can be used to extend the
range of the existing matrix.
| ||||
| emin | no | real | 0.3 | non-negative |
|
When creating a new response matrix (newrmf=yes), this is
the lower edge of the lowest incident-energy bin to be included.
When overwriting part of an existing matrix (newrmf=no),
this identifies the lower incident-energy bin in the interval to be
recomputed. This value will be lowered automatically as necessary
to maintain the contiguity of the bins, and can be used to extend
the range of the existing matrix.
| ||||
| rows | no | integer | 4000 | positive |
|
The number of incident-energy bins to cover the range from
emin to emax. Typically, when overwriting part of
an existing matrix (newrmf=no), the intention is to increase
the resolution of the matrix over a small interval. This can yield
a matrix of manageable size that has very high resolution in just
the right places.
| ||||
| fftdim | no | integer | 3 | 1-5 |
|
The wings of the various distributions that are convolved together
to form the narrow features of the line-spread function are
truncated to limit the size of the convolution space. An
increment of one in this parameter doubles the size of the
convolution space; the larger the convolution space, the slower
the computation and the less power is lost to the truncation.
| ||||
| newrmf | no | boolean | yes | |
|
Whether to create a new response matrix or edit an existing one.
| ||||
| The remaining parameters apply only when newrmf=yes. When editing an existing matrix these specifications are obtained from its contents. | ||||
| withspectrum | no | boolean | yes | |
|
Determines how various details of the response matrix are specified.
On true: enables parameter spectrumset, and the response
matrix is made compatible with this spectrum with respect to its
source, order, background correction, and channel definitions. On
false: enables parameters source, order, and
bkgcorrect, while the channel definitions are obtained from
the event list (parameter evlist). Note that a channel
rebinning factor cannot be specified except through the
spectrumset, an oversight which may be corrected in a future
version.
| ||||
| spectrumset | no | dataset | srspec.ds | |
|
A spectrum file produced by rgsspectrum
from the data in the
given event list (parameter evlist). Not modified. The
metadata from this file are used to ensure that the response matrix
produced is the suitable companion to this spectrum for the purpose
of analysis. A default filename cannot be inferred; when this
parameter is enabled (withspectrum=yes) an explicit value
must be provided, which then becomes the template from which other
filenames may be inferred.
| ||||
| source | no | integer | non-negative | |
|
If a spectrum file is not provided (withspectrum=no) this
parameter specifies the INDEX number of a source in the
source description list (parameter srclist), and thereby the
incidence angle and selection regions for the response matrix. The
default value, zero, indicates the SOURCEID attribute of the
EVENTS table in the event list (parameter evlist),
which is the source that was used to perform the aspect-drift
corrections.
| ||||
| order | no | integer | 1 | positive |
|
If a spectrum file is not provided (withspectrum=no) this
parameter specifies the reflection order for computing the response
matrix. In conjunction with the source parameter, this
identifies the energy selection region from the source description
list (parameter srclist). Although it does include the
contribution of the LSF for every order, the final response matrix
is nevertheless order-specific with respect to its effective area
and EBOUNDS extension.
| ||||
| bkgcorrect | no | boolean | yes | |
|
If a spectrum file is not provided (withspectrum=no) this
parameter specifies whether background correction should be included
in the response matrix computation. If yes, then the background
selection region from the source description list (parameter
srclist) is required. In HTR mode the background
correction to the response matrix is not defined, and this parameter
is quietly ignored.
| ||||
| withmirrorpsf | no | boolean | yes | |
|
Whether to convolve the standard mirror PSF distribution into the
LSF. This option should be used with care and knowledge. An
appropriate circumstance for omitting the mirror PSF is when a
custom angular distribution is provided (withangdist=yes)
that empirically establishes the PSF contribution.
| ||||
| withangdist | no | boolean | no | |
|
Whether to convolve a custom angular distribution into the LSF.
This option should be used with care and knowledge. Enables
parameter angdistset. Note that this distribution is
one-dimensional, an arbitrary function of dispersion off-axis angle.
It is rebinned onto the dispersion channel space according to the
order and central incident energy of each LSF. The problem of
reducing a two-dimensional image to this one-dimensional
distribution is left to the user.
| ||||
| angdistset | no | file | angdist.txt | |
|
An ASCII file containing a one-dimensional angular distribution
along the axis of dispersion in arc minutes. Not modified. Each
line of the file describes a distribution bin in terms of the start
and stop angles, and the value assigned to that bin. The
distribution may be normalized to any value, though unity is
expected and recommended. See the
input files
descriptions for further details.
| ||||
| dyneffarecorr | no | boolean | yes | |
|
Enable the dynamic effective area correction based on Chebyshev polynomials.
| ||||
| withrectification | no | boolean | no | |
|
Use empirical RGS effective area correction
| ||||
| witheffectiveareacorrection | no | boolean | no | |
|
Use RGS effective area correction
| ||||
| spectrumbinning | no | choice | lambda | lambda beta |
|
Dispersion binning type.
| ||||
| Parameter | Mand | Type | Default | Constraints |
