No cheby for phase

0 votes
While folding a baseband observation with dspsr, I sometimes obtain the following error:

dsp::UnloaderShare::unload error unloading division 0
Error::stack
        dsp::Archiver::unload
        dsp::Archiver::add Pulsar::Archive
        dsp::Archiver::set Pulsar::Archive
        dsp::Archiver::set Pulsar::Integration
        Tempo2::Predictor::iphase
Error::InvalidParam
Error::message
        no cheby for phase

Is there any chance that someone knows what it means?
asked Mar 31, 2017 by Caterina (200 points)

1 Answer

0 votes
To bypass the issue, I usually start the processing after N seconds from the beginning (option "-S" of dspsr). 1 is usually sufficient for me

However, this quick and dirty solution doesn't tell us anything about what's wrong with the beginning of the observation
answered Mar 31, 2017 by Caterina (200 points)
I have a hunch that this error may be due to rounding errors in the code that decides if a given absolute pulse phase (including integer turns) falls within the range of phases spanned by an individual set of Chebyshev polynomial coefficients.  I've checked in what I think should be a fix.   If you are able to reproduce the error, could you please check if it now works on your end?
...