[R-sig-Geo] Error in reading .shp

[Roger Bivand]

On Tue, 7 May 2013, Danielle Rappaport wrote:

> Hi there:
>
> I am trying to have R read a shapefile of a simple landscape classification
> (posted here <http://temp-share.com/show/3YgF8Kd4x>) and am getting an
> error message, which I'm not sure how to resolve:
>
>> Shape2 <-readShapePoly("JustForest_BahiaSul_1990.shp")
>
> Error in Polygon(coords = crds) : ring not closed
>
>
> I also tried specifying force_ring=TRUE, which does not seem to help:
>
>> shape3 <-readShapePoly("JustForest_BahiaSul_1990_Test.shp", force_ring=TRUE, delete_null_obj=TRUE)Error in Polygon(coords = crds) : ring not closed
>
> The same error results when I try with rgdal as an alternative:
>
>> shape1 <-readOGR("JustForest_BahiaSul_1990_Test.shp", layer="JustForest_BahiaSul_1990_Test")OGR data source with driver: ESRI Shapefile
> Source: "JustForest_BahiaSul_1990_Test.shp", layer:
> "JustForest_BahiaSul_1990_Test"
> with 19122 features and 6 fields
> Feature type: wkbPolygon with 2 dimensionsError in
> stopifnot(is.list(srl)) : ring not closed

Thanks for making the data available for reproduction, makes debugging 
much easier! The same happens in GRASS:

Counting polygons for 19122 features...
Importing map 19122 features...
  100%
-----------------------------------------------------
Building topology for vector map <test_tmp>...
Registering primitives...
v.in.ogr: index.c:159: RTreeInsertRect1: Assertion `r->boundary[i] <= 
r->boundary[3 + i]' failed.

So the shapefile is broken, and the problem is in the software used by 
your collaborator. The specific problem is that all of the x-coordinates 
in entity 15755, first ring are NaN, not finite.

If you are able to install sp and rgdal from R-forge from source checked 
out from SVN, you can get the corrected error message:

> Shape2 <-readOGR(".", "JustForest_BahiaSul_1990", useC=TRUE)
OGR data source with driver: ESRI Shapefile
Source: ".", layer: "JustForest_BahiaSul_1990"
with 19122 features and 6 fields
Feature type: wkbPolygon with 2 dimensions
Error in stopifnot(is.list(srl)) : non-finite x coordinate

and now by setting useC=FALSE, retrieve the entity number for errors. 
readOGR() will actually correct non-closed rings automatically, so the 
error message as it stood was misleading.

So your collaborator has to create a new file with finite coordinates.

Hope this clarifies,

Roger


>
>
> If it's of any use in troubleshooting, here is another shapefile
> (posted here <http://temp-share.com/show/KdPfyYg6h>) that was
> generated by the same collaborator of the same landscape just
> classified at another point in time, which R is reading without any
> problem.  I can't seem to figure out why
> "JustForest_BahiaSul_2007.shp" is being read successfully, and
> "JustForest_BahiaSul_1990.shp" is not.
>
> Many thanks in advance for offering your time and sage R-dom.
>
> Best, Danielle
>
>
> sessionInfo()
>
>
> R version 2.15.3 (2013-03-01)
> Platform: x86_64-w64-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United States.1252
> [2] LC_CTYPE=English_United States.1252
> [3] LC_MONETARY=English_United States.1252
> [4] LC_NUMERIC=C
> [5] LC_TIME=English_United States.1252
>
> attached base packages:
> [1] grid      stats     graphics  grDevices utils     datasets
> [7] methods   base
>
> other attached packages:
> [1] rgdal_0.8-8     maptools_0.8-23 lattice_0.20-13
> [4] sp_1.0-9        foreign_0.8-52
>
> loaded via a namespace (and not attached):
> [1] tools_2.15.3
>
>
>
>
> Danielle Ivonne Rappaport
> Master of Forestry 2012, Yale University
> Skype | Danielle_Rappaport
>
> 	[[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, NHH Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.
voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no