[ncl-talk] Sub: Regrid error..

[Mary Haley]

I am still looking into the seg fault error with the ESMF group. It turns
out that the Intel compiler is doing something under the hood that causes
numerical problems, and we are having to try different compiler options to
fix the problem. Still no luck, but we think we're close.

Meanwhile, since you are setting:

    Opt at SrcGridLat        = P1&lat       ; source grid
    Opt at SrcGridLon        =P1&lon

You do not also want to set this, as it might confuse the issue:

 Opt at SrcGridType    = "0.93deg"

If this doesn't fix the problem, or using a GNU version of NCL/ESMF doesn't
work either, then please email ncl-talk at ucar.edu and let us know.  I may
email you offline to ask if I can have access to your files.

Are your variables large by any chance?  That is, what does the following
output produce?

  printVarSummary(P1)
  printVarSummary(p)

--Mary


On Mon, Nov 6, 2017 at 3:26 PM, Laura Fowler <laura at ucar.edu> wrote:

> Hi Dale:
>
> I encountered the same kind of segmentation fault when using ncl/6.4.0
> compiled with the intel compiler. I addressed this issue to ncl-talk
> on October 25th. Note that my settings were different and I was not
> using Opt at DstRegional = True.
>
> Mary Haley suggested to use ncl/6.4.0 compiled with the gnu compiler.
> Changing compilers solved the issue I had with the "segmentation
> fault". I am forwarding you her reply separately. You may be
> encountering the same issue too. Hope this helps.
>
> Cheers,
> Laura
>
>
>
>
> On Mon, Nov 6, 2017 at 2:59 PM, dale zuri <dalezuri at gmail.com> wrote:
> > Hi NCL community,
> > Could someone help me understand segmentation fault error?
> >
> > Thanks
> > DZ
> > begin
> > ;;;;;;;;;;;;;;;;;ERA;;;;;;;;;;;;;;;;;;;;;;;;;;;
> >   iz = addfile("psa_mask_cfsv2.nc","r")
> > P1=iz->mask
> >   ds = addfile("nlwrs.sfc.gauss.1948.nc","r")
> >   p = ds->nlwrs
> >   printVarSummary(p)
> > ;****************************************************
> > ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
> > ;--Set up regridding options
> >     Opt                   = True
> >
> > ;---"bilinear" is the default. "patch" and "conserve" are other options.
> > ;    Opt at InterpMethod      = "neareststod"        ;;---Change (maybe)
> >
> >
> >     Opt at SrcGridLat        = P1&lat       ; source grid
> >     Opt at SrcGridLon        =P1&lon
> >  Opt at SrcGridType    = "0.93deg"
> >   Opt at DstGridLat        = p&lat           ; destination grid
> >     Opt at DstGridLon        =p&lon
> >     Opt at DstRegional       = True              ;;--Change (maybe)
> >
> >     Opt at ForceOverwrite    = True
> >     Opt at PrintTimings      = True
> >     Opt at Debug             = True
> >
> >
> >     P = ESMF_regrid(p,Opt)     ; Do the regridding
> >
> > end
> >
> > ERROR I get:
> >
> > (0)    ESMF_regrid_with_weights: retrieving interpolation weights ...
> > (0)    ESMF_regrid_with_weights: calling sparse_matrix_mult to apply
> > weights...
> > Segmentation fault (core dumped)
> >
> > _______________________________________________
> > ncl-talk mailing list
> > ncl-talk at ucar.edu
> > List instructions, subscriber options, unsubscribe:
> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
> >
>
>
>
> --
> !-----------------------------------------------------------
> --------------------------------------------------
> Laura D. Fowler
> Mesoscale and Microscale Meteorology Division (MMM)
> National Center for Atmospheric Research
> P.O. Box 3000, Boulder CO 80307-3000
>
> e-mail: laura at ucar.edu
> phone: 303-497-1628
>
> !-----------------------------------------------------------
> --------------------------------------------------
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20171108/37491b8b/attachment.html>