It does not matter for the netCDF file.
In certain cases the binary files may not be portable between compilers.
But generally they will conform (if using relatively new compilers).

So:
1) For netCDF it does not matter
2) For the binary files of siesta, it may matter, but test and most likely
it will work! :)

2016-05-03 16:34 GMT+02:00 Joscha Hekele <[email protected]>:

> Dear Thomas,
>
> thanks for your reply. So my TS compilation doesn't support netCDF as it
> should, I guess. I will try to recompile it.
> By the way: does it matter if TS+netCDF are compiled with different
> compilers (intel/gcc) than Inelastica[Eigenchannels] components (same
> versions of netCDF, python, numpy, etc. given)? I suspect that this doesn't
> influence the interfacing file format?
>
> Thanks and best regards,
> JH
>
> Thomas Frederiksen schrieb am 2016-05-02:
> > Dear Joscha,
>
> > >From the error message below it appears that the required files *.
> ion.nc are missing from your TranSIESTA run.
>
> > Best regards,
> > Thomas Frederiksen
>
>
> > On 2 May 2016 19:14:33 CEST, Joscha Hekele <[email protected]>
> wrote:
> > >Dear users,
> > >
> > >I'm having many problems with getting to run Inelastica. I rigorously
> > >followed Guanping Zhang's guide (using only the software versions given
> > >there) but it just wouldn't work (with self-compiled ATLAS,
> > >libquadmath.so.0 is missing).
> > >So I tried to install it by using the software provided on our cluster
> > >(python 2.7.9 with numpy 1.9.1). I used ifort 15.0.1.133 to compile the
> > >other packages. These were netcdf-3.6.3 and ScientificPython-2.8.1.
> > >Transiesta was compiled with GCC and the same netcdf package.
> > >Importing Inelastica and the provided scripts seems to work but when I
> > >try to run EigenChannels, I get the following error:
> > >
> > >Traceback (most recent call last):
> > >File
> > >"/home/j/j_heke01/Software/inelastica-code-407_intel/bin/EigenChannels",
> > >line 12, in <module>
> > >    Module.main(options)
> > >File
> >
> >"/home/j/j_heke01/Software/inelastica-code-407_intel/lib/python2.7/site-packages/Inelastica/EigenChannels.py",
> > >line 143, in main
> > >    options.DeviceAtoms[1]+L,lasto)
> > >File
> >
> >"/home/j/j_heke01/Software/inelastica-code-407_intel/lib/python2.7/site-packages/Inelastica/SiestaIO.py",
> > >line 1312, in BuildBasis
> > >    ions[int(CSL[i][0])] = ReadIonNCFile(head+'/%s.ion.nc'%CSL[i][2])
> > >File
> >
> >"/home/j/j_heke01/Software/inelastica-code-407_intel/lib/python2.7/site-packages/Inelastica/SiestaIO.py",
> > >line 1263, in ReadIonNCFile
> > >    file = nc.NetCDFFile(filename,'r')
> > >File
> >
> >"/home/j/j_heke01/Software/ScientificPython-2.8.1_intel/lib/python2.7/site-packages/Scientific/IO/NetCDF.py",
> > >line 177, in NetCDFFile
> > >    return apply(_NetCDFFile, args)
> > >IOError: netcdf: No such file or directory
> > >
> > >I recompiled ScientificPython with our cluster-provided netcdf (version
> > >4.1.2) but I still get the same error message. The environment seems to
> > >be set up appropriate too.
> > >Can anybody help me?
> > >
> > >Best regards,
> > >JH
>



-- 
Kind regards Nick

Responder a