SAS watchout page

SAS watchout page

In this page some issues concerning SAS and XMM-Newton data analysis are reported, along with recommended workarounds and/or solutions, and useful tricks and tips.

Many of the items below are extracted from the SAS HelpDesk questions database, and frequently updated. If a SAS topic of interest is not reported in this page, users are invited to search in the HelpDesk database before contacting the XMM-Newton HelpDesk.

Actual SAS Version related issues

SAS 11.0.0

• GTI filtering in emosaicproc: GTI parameters (pngtilists, mos1gtilist and mos2gtilist) are not used for any filtering. Is there any workaround for such a problem? -- Answer
• Why won't emosaic combine my image masks? -- Answer
• Why do I need to update ESAS CalDB? -- Answer
• When running the SAS 11.0.0 GUI (sas) for Mac OS X Leopard on PowerPC, fonts appear illegible. How to fix the problem? -- Answer
• The html documentation available on both, on-line and the public distribution, is missing the links to some well known tasks. For example the links to epproc and emproc are missing in the list of all tasks. However, the pdf documentation distributed publicly includes the specific documents on these tasks. Is there any fix for such a problem? -- Answer

SAS 10.0.0

• Why do I get problems ingesting RGS (pipeline) spectra into XSPEC V12.x and how can I solve this? Answers
• HEASOFT / SAS: tips and trick -- Info
• Can I avoid the strange error from FTOOLS fcalc when running emchain? -- Answer
• Should I update my CCF now that I do have installed SAS 10.0? -- Answer

SAS 9.0.0

• Should I update SAS to SAS 10.0? -- Info
• pn_back not available under esas -- Info
• Problems to run epspllitter on non-imaging mode PN observations? Please check -- here how to solve it.
• Have you installed SAS 9 on Mac OS X Leopard before 30 June? Please update a library -- Info

SAS 8.0.0

• Performing absolute timing analysis of data taken in 2009? Please update a file -- Info
• MacOS X Leopard: the syndrome of the ghost file -- Info
• On MacOs versions, tasks get sometimes kill signals from system -- Info
• Some problems with the new task epiclccorr -- Info

Evergreen tips and tricks

• Should I use evselect or epiclcorr for obtaining EPIC light curves? -- Answer
• Should I care about the binning of EPIC spectra? -- Answer
• How can I correct for time sensitivity degradation of the OM detector? -- Answer
• These long parameter names are really annoying ... -- Answer
• How do I compare two Calibration Index Files? -- Answer
• How do I calculate the BACKSCAL keyword for my EPIC spectra? -- Answer
• Can I generate EPIC response matrices with all types of spectral binning? -- Answer
• Are there any differences between "standard" and SAS-generated EPIC response matrices? -- Answer
• Can I estimate the pile-up level in my observation? -- Answer
• Can I perform the barycentric correction on OM data in Fast Mode? -- Answer
• How can I convert from OM count rates to fluxes? An update -- Answer
• My grace window seems not to start. Is there anything wrong with the SAS? -- Answer
• Rectangular grace windows: how to get them? -- Answer
• Dealing with MOS1 data in Timing Mode: I heard that one column in MOS1 Timing Mode is affected by "high offset", can I still analyze the data? -- Answer

ODF

• Problems with Time Correlation files in period Oct.2007-July 2008 -- Info
• Why does it take so long to produce an ODF? -- Answer

PPS products

• Why do I get problems ingesting RGS (pipeline) spectra into XSPEC V12.x and how can I solve this? Answers
• How do I know which CCF were used to reduce my event list? -- Answer
• I do not find EPIC (or OM) processed data in my PIPEPROD directory. Did something go wrong with the pipeline processing? -- Answer

SciSim

• Is the outcome of SciSim (V4.0) really a realistic representation of science data obtained with XMM-Newon? -- Answer

Previous SAS Version related issues

SAS 7.1.2

• SAS GUI commands do not run on Mac OS X Leopard, PowerPC and Intel. How to make them work? -- Answer

SAS 7.1.0

• When SAS is initialized first and then HEASOFT/HEADAS, SAS programs do not work anymore. Why? -- Answer
• Some tasks which produce output in postscript, do not work and give an error. Why? -- Answer
• rgsproc: error (ParamNotInRange), Parameter 'entrystage' or rgsproc: error (ParamNotInRange), Parameter 'finalstage' -- Answer
• No labels and tick marks are shown up on plots made from some SAS tasks. What is the problem and how to fix it? -- Answer
• I have built HEADAS 6.3 from source code and I am initializing it before SAS. Now when I run any HEADAS command, it fails with the error ...xmmsas_20070708_1801/libextra/libgcc_s.so.1: version `GCC_4.2.0' not found (required by /usr/lib/libstdc++.so.6) . What is the problem and what can I do? -- Answer

SAS 7.0.0

• Is it true that lccorr is unreliable? -- Answer
• Artifacts in XMM-Newton time series analysis? -- Answer
• There is no labelling of the axes in the rgsimplot output. Why? -- Answer
• Why is the BACKSCAL value wrong in TIMING mode? -- Answer
• On-line documentation on rgsproc and rgsregions parameters xpsfincl, xpsfexcl and pdistinc default values differs from their tasks default values. What is correct? -- Answer
• On RHEL 4 update 4, omichain (and omfchain) fail to run properly due to a changed version of ghostscript. -- Answer

SAS 6.5.0

• I cannot filter on the TIME an RGS event file produced with a former SAS version. What am I doing wrong? -- Answer
• Why does "atthkgen" fail due to "time correlation" related problems, working with the MacOS version of SAS? -- Answer
• Quadrant in PN with fewer events than expected? ? -- Answer
• Is it safe to use SAS 6.5 under a Mac OSX 10.4 (Tiger)? -- Answer
• What is about the patch to SAS 6.5? Do I need new calibration files? -- Answer
• Why do I get the error "UnsupportedAlgoId" when processing MOS data? -- Answer
• Can I use my old scripts in SASv6.5, which include calls to the very much refurbished emproc / epproc? -- Answer
• I get a segmentation fault when trying to run rgssuperrmf. Am I doing something wrong? -- Answer

SAS 6.1.0

• Why do I get bad MOS spectral fits now using SAS 6.0 and SAS 6.1? -- Answer
• Can I use SASv6.1 with any processed event lists? -- Answer
• Reprocessing pn data with SASv6.1 yields event lists with less photons than with SASv5.4.1. Am I doing something wrong? -- Answer
• Bug in tabgtigen -- Answer
• My BACKSCAL (or effective area) has crazy values: help! -- Answer
• Can I use arfgen for spectra extracted from regions of arbitrary shape? -- Answer
• I get the error: "Subscript 1 of PATTFRACTION (value 1) is out of range (1:0)" with rmfgen -- Answer
• My exposure maps are identically null everywhere. Am I doing something wrong? -- Answer
• How far can I trust the results of my timing analysis? -- Answer
• The OM chains fail saying that an OM exposure mode is "UNDEFINED" in the ODF. However, ... -- Answer

Old SAS versions

6.0.0

• I have a problem running rgsproc, because a library is missing. How can I solve it? -- Answer
• My SunOS workstation cannot run SAS v6.0 due to ``missing libraries'' ... -- Answer
• If I close (Exit) the xmmselect GUI, the X display is suddenly restarted. What can I do? -- Answer
• Can I run old SAS scripts with the new SAS v6.0, given its new parameter interface? -- Answer
• Shall I update my calibration files with SASv6.0? -- Answer
• My odfingest run complaints about TimeOutOfRange. Is my ODF healthy? -- Answer
• Are there any software requirements on the analysis of MOS/RGS post-cooling data? -- Answer
• I get an error when processing my pn data with epframes. Any clues? -- Answer
• The coordinates of regions I import in xmmselect from ds9 are fully out-of-bounds. Am I doing something wrong? -- Answer
• The emission line energies in my MOS spectrum seem a bit too low. Any calibration issues? -- Answer
• I observe a shift in time between OM and other light curves. -- Answer

5.4.1

• I do not find a SAS v5.4.1 version for Red Hat 7.2 -- Answer
• Running SAS v5.4.1 for Red Hat 7.2, some programs crash mysteriously. Anything wrong with the SAS? -- Answer
• My SAS seems not to recognize my Red Hat Perl version. Is there something I can do to tame it? -- Answer
• I have Red Hat 8.0 installed on my workstation. Can I use the SAS? -- Answer
• Does BACKSCAL take into account all the "blank areas" in a spectrum extraction region? -- Answer
• Can I use evigweight on all types of event lists? -- Answer

5.3.0

• A library error prevents me from running any SAS tasks. What have I done wrong? -- Answer
• I get a FITSIO or DAL error when I try to read an event list file. What's happening? -- Answer
• My MOS data get strangely screened by emproc. Am I doing something wrong? -- Answer
• I am unable to analyze RGS products, created with SAS v5.0, using later versions of SAS. Is there a workaround? -- Answer
• Are there any incompatibilities between CCF constituents and SASv5.3.0? -- Answer
• My SAS seems to be inconsistent with a MOS redistribution CCF. What can I do? -- Answer
• LOW, MEDIUM, HIGH, EXTENDED ... how do I get "the" EPIC PSF from calview? -- Answer
• ONTIME is much shorter than the elapsed time. Am I doing something wrong? -- Answer
• xmmselect does not seem to recognize the screening criteria used to generate a filtered event list -- Answer
• I get strange edge-like features in my pn spectrum. Is therefore something I need to know? -- Answer

5.2.0

• SAS seems not to recognize the ASCII summary file pointed by SAS_ODF. Am I doing something wrong? -- Answer
• Good start! odfingest (or cifbuild) complaints that the summary file does not have the CONFIGURATION record ... -- Answer
• Extracting a scientific product with xmmselect/evselect I get a warning concerning the exposure time. Should I worry? -- Answer
• The circular annulus region in ds9: annulus(x,y,rinner,router) gets imported by xmmselect as annulus(x,y,0,rinner). What is wrong? -- Answer
• rgsproc crashed with an out-of-range error on betabinning. Is it a bug? -- Answer