[R-sig-Geo] gDistance problem on RHEL 6 64-bit

[Roger Bivand]

On Thu, 5 Feb 2015, Hollister, Jeff wrote:

> Roger,
>
> Thanks for getting back to me.  I thought it was a GEOS problem too.  I 
> have updated it and re-installed rgeos.
>
> Curret GEOS is:
>> version_GEOS()
> [1] "3.4.2-CAPI-1.8.2 r3921"
>
> I tried several of the other rgeos examples (gBuffer, gWithin, 
> gEnvelope, gUnion, gLength ...) and they all worked.  gDistance bombs 
> the same as before, as does gWithinDistance.
>
> My version of gcc appears to be a bit old:
>
> -bash-4.1$ gcc --version
> gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-11)
>
> Could that be the issue?

No, since the other examples work, it isn't likely. On my former RHEL5, 
gcc is 4.1.2 of 2008 and still works.

If you need to follow this up, that is that your RHEL6 machine is mission 
critical, make sure that it is fully updated, remove and reinstall GEOS, 
run /sbin/ldconfig, look for other GEOS installations, remove all GEOS, 
reinstall GEOS; once you are sure that GEOS is installed uniquely, and 
runs its own make check correctly, remove rgeos and reinstall it. If the 
problem persists, run R under GDB (R -d gdb) to give a traceback on error, 
and run:

.Call("rgeos_distance", rgeos:::.RGEOS_HANDLE, pt2, pt1, c(FALSE, FALSE),
   PACKAGE = "rgeos")

as the most stripped-down version. What is of interest is where in 
src/rgeos_misc.c in rgeos_distancefunc() the failure occurs, that is, does 
it occur in converting the sp objects into GEOS representation, in 
computing the distance, or in destroying the GEOS representations of the 
input objects?

Does gDistance(pt2, pt1, byid=TRUE) make a difference (probably not)?

Roger


>
> Thanks again for your help!
> Cheers,
> Jeff
>
>
>
> *****************************
> Dr. Jeffrey W. Hollister
> Research Ecologist
> 27 Tarzwell Drive
> Narragansett, RI 02879
> (o) 401 782 9655
> hollister.jeff at epa.gov
> Personal Site
> *****************************
>
> -----Original Message-----
> From: Roger Bivand [mailto:Roger.Bivand at nhh.no]
> Sent: Thursday, February 05, 2015 3:28 PM
> To: Hollister, Jeff
> Cc: r-sig-geo at r-project.org
> Subject: Re: [R-sig-Geo] gDistance problem on RHEL 6 64-bit
>
> On Thu, 5 Feb 2015, Hollister, Jeff wrote:
>
>> Hello all,
>>
>> I am having a problem with rgeos::gDistance on RHEL 6.  Everytime I
>> try to get a point to point distance, I get
>>
>> R: GeometryComponentFilter.cpp:34: virtual void 
>> geos::geom::GeometryComponentFilter::filter_ro(const 
>> geos::geom::Geometry*): Assertion `0' failed. Aborted (core dumped)
>>
>> This only occurs on my RHEL 6 machine.  Win 7 and Ubuntu 14.04 work fine.
>
> The key question is how you installed rgeos and GEOS on these platforms.
> Most likely rgeos on Windows 7 is the CRAN binary statically linked to GEOS, and on RHEL6 and Ubuntu 14.04 rgeos is installed from source and links to the GEOS found on those platforms. We need the output of:
>
> version_GEOS()
>
> on each platform. Most likely you've not installed GEOS from source on either of the two Linux platforms. The cause is most likely a mismatch on the RHEL platform, possibly that GEOS has fallen behind C++, or that GEOS got updated without you re-installing rgeos. Is gDistance() the only function that doesn't work - can you run example(gDistance) and/or the examples of other functions?
>
> Roger
>
>>
>> Some code to reproduce the problem:
>> library(sp)
>> library(rgeos)
>> pt1<-SpatialPoints(data.frame(x=c(1,3),y=c(3,1)))
>> pt2<-SpatialPoints(data.frame(x=1,y=1))
>> gDistance(pt2,pt1)
>>
>> More info:
>>> sessionInfo()
>> R version 3.1.2 (2014-10-31)
>> Platform: x86_64-redhat-linux-gnu (64-bit)
>>
>> locale:
>> [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
>> [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
>> [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
>> [7] LC_PAPER=en_US.UTF-8       LC_NAME=C
>> [9] LC_ADDRESS=C               LC_TELEPHONE=C
>> [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
>>
>> attached base packages:
>> [1] stats     graphics  grDevices utils     datasets  methods   base
>>
>> other attached packages:
>> [1] rgeos_0.3-8 sp_1.0-15
>>
>> loaded via a namespace (and not attached):
>> [1] grid_3.1.2      lattice_0.20-29
>>
>> Thanks,
>> Jeff Hollister
>>
>> *****************************
>> Dr. Jeffrey W. Hollister
>> Research Ecologist
>> 27 Tarzwell Drive
>> Narragansett, RI 02879
>> (o) 401 782 9655
>> hollister.jeff at epa.gov<mailto:hollister.jeff at epa.gov>
>> Personal Site<http://jwhollister.com/>
>> *****************************
>>
>>
>> 	[[alternative HTML version deleted]]
>>
>> _______________________________________________
>> R-sig-Geo mailing list
>> R-sig-Geo at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>
>
> --
> Roger Bivand
> Department of Economics, Norwegian School of Economics, Helleveien 30, N-5045 Bergen, Norway.
> voice: +47 55 95 93 55; fax +47 55 95 91 00
> e-mail: Roger.Bivand at nhh.no
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

-- 
Roger Bivand
Department of Economics, Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.
voice: +47 55 95 93 55; fax +47 55 95 91 00
e-mail: Roger.Bivand at nhh.no