[Date Prev][Date Next][Thread Prev][Thread Next][Search] [Main Index] [Thread Index] [HEASARC Archives]

A bug in correctrdd

A Bug in Correctrdd

We have just discovered a bug in FTOOLS v4.1 version of correctrdd.
When combined with sisrmg, the bug results in a response matrix that
is appropriate for non-RDD-corrected data.  We have fixed the problem,
and the corrected version will be included in the forthcoming FTOOLS
v4.2 release.

[1] How to check if you are affected:

RDD-corrected event files, and spectral files extracted from them,
should contain the keyword RDDCOR_V ("Version of RDDcorrect
algorithm used") in the primary header.  The correct value is 1;
the bug was to write it out as a Logical.  If you see

RDDCOR_V=                    T / Version of RDDcorrect algorithm used

then that file is affected by the bug.

The response matrix made with an incorrect pha file will have

RQRDCV  =                    0 / Version of RDD correction s/w applied to data

(0 is the correct value for non-RDD corrected data)

[2] Workaround

You can use 'fv' or 'fmodhead' to change the primary header keyword
in your spectral file to

RDDCOR_V=                    1 / Version of RDDcorrect algorithm used

Re-generate the response using sisrmg, with the corrected spectral
file as the input.  You should check that the RQRDCV keyword in
the RMF file now has the value 1.  If you intend to keep the RDD-corrected
event file for further analysis, we recommend you fix the event file.

Note, however, that the response for RDD-corrected data are currently
poorly calibrated.

We apologize for any inconveniences.