RadxConvert fails with cfradial files written using PyART


I am trying to convert some cfradial files (non-NEXRAD) to dorade format. However, RadxConvert throws the following error message:

INFO - RadxConvert::Run
Input path: data/ARMOR_qc1.nc
ERROR - RadxConvert::_readFile
  path: data/ARMOR_qc1.nc
ERROR - NcfRadxFile::_readPositionVariables
  latitude is incorrect type: nc3Float
  expecting type: double
ERROR - NcfRadxFile::readFromPath::_readPath
ERROR - GenericRadxFile::readFromPath
  File format not recognized: data/ARMOR_qc1.nc

System configuration:

MacOS High Sierra v 10.13.6, and

brew info lrose-core                                                        
lrose-core: stable core-20190130
/usr/local/Cellar/lrose-core/core-20190130 (2,400 files, 98.1MB) *
  Built from source on 2019-03-26 at 16:30:15
From: /usr/local/opt/lrose-core/.brew/lrose-core.rb
==> Dependencies
Required: hdf5 --enable-cxx ✘, netcdf --enable-cxx-compat ✘, udunits ✘, fftw ✘, flex ✘, jpeg ✔, libpng ✔, qt ✘, szip ✔, pkg-config ✔
==> Requirements
Required: x11 ✔

Hi Milind,

Can you share the command used to run RadxConvert and/or what parameters you modified in the parameter file?

I just ran the following command on some other ARMOR data you’d sent me previously and it does appear to have created sweep files.
RadxConvert -dorade -f ARMOR_20160401013612_qc1.nc

I will note that my machine is using the May 2021 release of LROSE on Catalina.


Hi Jen,

Thanks for the reply. Yes, I am also able to convert the original cfrad files (including the one you mentioned above). However, there seems to be an issue with the files I generated upon some post-processing using PyART. I have emailed you one of those post-processed files. Could you please try converting that one?

Hi Milind,

I assume that you’ve now been able to update LROSE to a more recent version, you’ll be testing this conversion out again. Let me know if you still encounter the problem or if it’s been fixed.


Hi Jen,

Yes, the new version (lrose-core-20211114) works fine and shows only a warning instead of an error.

In case anyone else is interested, here is how I installed it on MacOS Catalina: