EUV Package - detpos not finding detector.tab
DamianJC wrote on Jun 23, 2015
Hi All,
This may be obscure and outdated, but I am trying to make some EUVE light curves with the
EUV package (v 1.9 from 2/1/00 on a mac with IRAF 2.16) and one of the functions, detpos (detector position)
is not finding the reference data, even though it is in the directory.
PACKAGE = euvtools
TASK = detpos
input = events.fit[orientation] Name of input aspect ST table
output = test Prefix for output aspect ST tables
(refdata= reference/detector.tab) Name of ST table for reference data
(detecto= ds) Detector list [sca|scb|scc|ds]
(ra = 51.5) Source right ascension
(dec = 27.5) Source declination in decimal degrees
(raunits= degrees) Units of RA [degrees|hours]
(time = TIME) Time column of input aspect table
(aspectw= ASPECTW) Aspect W input column name
(aspectx= ASPECTX) Aspect X input column name
(aspecty= ASPECTY) Aspect Y input column name
(aspectz= ASPECTZ) Aspect Z input column name
(mode = al)
ERROR: Unable to find table, reference/detector.tab
even though it is in the directory:
euvtools> ls reference/detector.tab
reference/detector.tab
and I have tried multiple paths to it. At first I thought detector.tab had some problems, but it seems now
detpos is hardwired to look for the reference file in the wrong place.
any help is greatly appreciated.
-Damian
Mike Fitzpatrick wrote on Jun 24, 2015
The EUV package has been frozen for quite a while and was not updated for 64-bit systems, so you're still running the original 32-bit binaries. The first thing I'd check is whether the detector.tab file can be read at all, e.g. can you do
cl> tinfo reference/detector.tab
cl> count reference/detector.tab
The first command would verify the table file itself can still be read (maybe there are byte-swapping issues?), the second would verify the file can be opened as a file. The latter test is because some 32-bit binaries cannot read files on 64-bit filesystems and I want to be sure this isn't happening here (if it is, try copying the detector.tab file to /tmp and using it from there). Otherwise, what version of the EGODATA files are you using? Have you tried accessing the FITS files directly rather than using the converted .tab versions?
DamianJC wrote on Jun 26, 2015
Hi Fitz,
thanks for looking at this. Here are the results to your suggested commands:
euvred> tinfo reference/detector.tab
# reference/detector.tab
7 rows written to table
63 columns defined
1 header parameters written to table
94.0 row length in units of SZ_REAL
94.0 amount of row length used
6 records allocated for header parameters
63 space allocated for column descriptors
table type: stsdas
euvred> count reference/detector.tab
0 1 34 reference/detector.tab
The sys admin here added some write statements and we found that the detpos routine
was not finding ds_distort.fits file that is called from detector.tab. There were several versions of ds_distort in the reference dir (.fits and .hhd formats). once the ds_distort.hhd file was removed, detpos worked.
I have now moved on to another error, where the xray.xtiming package does not read the time filters added for the euv package.
add the time filter with qphedit:
qpoe = "ds.qp" qpoe(s) to be edited
fields = "deffilt" fields to be edited
value = "time=(gt_ds.ta)" value expression
We get the error:
ERROR: QPEX: bad range specifier in event attribute filter
"qp_mkindex=no, qp_key="NONE", qp_debug=debug)"
line 124: xtiming$timsort.cl
called as: `timsort ()'
But if I do not add the time filter with qphedit, then timsort works. this time filtering used to work.I know the X-ray package can be very picky for formats and will look at my old notes, but hoping you have seen this before. The euv routine qpmkim also will not work with the time filters.
thanks much!
-Damian
Last post on Jun 26, 2015