When processing continuous observation of more than 1 day, an error occurred from DSPSR:
Error::message
epoch 60230.28412758417979894210 not spanned by ChebyModelSet
The error is related to TEMPO2 wrapper T2Predictor
, where TEMPO2 complains ChebyModelSet_OutOfRange
.
This points to the command line of DSPSR calling TEMPO2, which does have a span of one day:
tempo2 -npsr 1 -f pulsar.par -pred "fast 60229.283869345 60230.283869345 200.006942749023438 350.006942749023438 12 2 3599.9999999999998" > stdout.txt 2> stderr.txt
Further searching shows the time span comes from dspsr/Signal/Pulsar/Fold.C
, line 243.
Changed this magic number
diff --git a/Signal/Pulsar/Fold.C b/Signal/Pulsar/Fold.C
index 009c1615..0c5c42f9 100644
--- a/Signal/Pulsar/Fold.C
+++ b/Signal/Pulsar/Fold.C
@@ -262,7 +262,7 @@ dsp::Fold::get_folding_predictor (const Pulsar::Parameters* params,
/*
* Tempo2 predictor code:
*
* Here we make a predictor valid for the next 24 hours
* I consider this to be a bit of a hack, since theoreticaly
* observations could be longer, and it's a bit silly to make
* such a polyco for a 10 min obs.
*
*/
- MJD endtime = time + 86400;
+ MJD endtime = time + 10*86400;
generator->set_site( observation->get_telescope() );
generator->set_parameters( params );
generator->set_time_span( time, endtime );
so folding can be done further.