search

lofar-astron / factor

Facet calibration for LOFAR
http://www.astron.nl/citt/facet-doc
GNU General Public License v2.0
19 stars 12 forks source link

Python argument types in ParmDB._getValues did not match C++ signature #225

Closed tikk3r closed 5 years ago

tikk3r commented 5 years ago

Trying some further out sources (with international station) the pipeline crashes at convert_merged_selfcal_parmdbs with the following error:

2018-10-24 21:09:22 ERROR   node.lofar7.strw.leidenuniv.nl.python_plugin: Python argument types in
    ParmDB._getValues(parmdb, str, numpy.ndarray, numpy.ndarray, numpy.ndarray, numpy.ndarray, bool, bool)
did not match C++ signature:
    _getValues(LOFAR::BBS::PyParmDB {lvalue}, std::string parmnamepattern, double sfreq, double efreq, double freqstep, double stime, double etime, double timestep, bool asStartEnd=True, bool includeDefaults=False)
    _getValues(LOFAR::BBS::PyParmDB {lvalue}, std::string parmnamepattern, double sfreq=-1e+30, double efreq=1e+30, double stime=-1e+30, double etime=1e+30, bool asStartEnd=True, bool includeDefaults=False)
2018-10-24 21:09:22 INFO    node.lofar7.strw.leidenuniv.nl.python_plugin: Total time 1.7829s; user time: 0.9904s; system time: 0.4249s

I'm using v3_2_2 of the LOFAR software and commit daf1a09 of factor (October 1st). Two other sources went fine. One is at the phase center and one is about 0.5 degree away. This error shows up for a couple of sources about 1.5 degrees away.

Is this related to my software install or did something go wrong during calibration? It made the 02-image.fits through 42-image.fits and they all contain the source.

darafferty commented 5 years ago

I'm not sure, but I suspect it has something to with the frequency and time grids being wrong somehow. Can you make available the .merge_phase_parmdbs file, the .smooth_amp2 file, and the file listed in facetselfcal/facet_patch_XX/mapfiles/expand_preapply_parmdbs.mapfile if this mapfile exists? I can then try it myself and see if I can reproduce the problem.

tikk3r commented 5 years ago

The expand_preapply_parmdbs.mapfile does not exist there. I have uploaded the parmdbs here: https://owncloud.strw.leidenuniv.nl/index.php/s/C35n47OLmvrObxe

darafferty commented 5 years ago

Thanks -- I think I fixed the problem (by reading the *. smooth_amp2 file in smaller blocks -- likely there is a bug in lofar.parmdb that limits how much can be read in some cases). Can you update and retry (you should be able to simply restart Factor without resetting anything)?

tikk3r commented 5 years ago

Yep, that looks like it fixes it. It's now at the prepare_imaging_data step. Thanks again.

tikk3r commented 5 years ago

It finished the facet perfectly, so it worked.