Hi Matt,
I tried to reproduce the error starting with the raw data files for that
observation. I didn't get to the phase correction step because your
fts2init program segfaulted. Did that step work successfully for you?
At first it segfaulted because fts2_validatemirrorpositions set *nStop to
-1, and after I fixed that it segfaults in the call to smf_write_smfData
on line 370. The first sounds like the problem Malcom mentioned that Josh
was going to work on before leaving, and the second I didn't investigate
yet. So I was wondering if you have any patches which haven't been pushed
to github?
Best regards,
Graham
> Hi Tim,
>
> Here is a link to a compressed version of a sample file of the type in
> question: http://www.filedropper.com/s8c20120925000260003ftssdf if you
> would care to try it.
>
> Thanks,
> Matt
>
> On 2012-12-06, at 3:16 PM, Tim Jenness wrote:
>
> We don't know what this "type" of file is so I'm not sure. If you give
> us an example then we can help. I'm guessing a previous step has done an
> NDF section into the data array but has not adjusted the dimensionality
> of JCMTSTATE. SMURF is meant to be able to attempt to correct for this
> by offsetting into the JCMTSTATE array. If we have an example file we
> might be able to work out what changed. smf_open_file only changed a
> little in January. sc2store changed in September but that was a minor
> tweak to NDG.
>
>
> On Thu, Dec 6, 2012 at 9:57 AM, Sherwood, Matt <[log in to unmask]<mailto:[log in to unmask]>> wrote:
> Hi David,
>
> It is true that the s8c20120925_00026_0003_phs.sdf does not exist. The
> reason is that this is the intended output of the smurf recipe task
> _FTS2_APPLY_PHASE_CORRECTIONDS_, which fails when attempting to open
> s8c20120925_00026_0003_fts.sdf. This file is emitted from the previous
> step (_FTS2_INIT_). This error occurs in smurf_fts2_phasecoorrds.c upon
> calling smf_open_file().
>
> This is the reason for my inquiry. What is now causing smf_open_file to
> fail with this type of file as input? My predecessor Josh states that
> this step worked recently, but now does not.
>
> Thanks,
> Matt
----
Starlink User Support list
For list configuration, including subscribing to and unsubscribing from the list, see
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A0=STARLINK
|