Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

How can I handle an "orphaned hole" in a polygon shapefile in R or ArcGIS?

I am trying to make a spatial overlay between two polygons ("grd" and "range") using the over function (rgeos).

This code works for all files but one:

range.grd.overlay <- grd[!is.na(over(grd, range)), ]@data$grdID # get gridcells that overlap with range

and get this error message:

Error in createPolygonsComment(p) : 
  rgeos_PolyCreateComment: orphaned hole, cannot find containing polygon for hole at index 9

I understand that there is a hole without a surrounding polygon somewhere in the shapefile, but I can't figure out how to work around it. Can I simply delete the hole at index 9 ? How ? If you have a solution to fix the shapefile in ArcGIS that would also be fine for me.

Thanks a lot in advance for your help !!

Regards, Carsten

like image 338
user1012744 Avatar asked Jul 11 '12 12:07

user1012744


People also ask

How do I fill a polygon in Arcmap?

Select the polygons that have holes you want to fill. on the Production Editing Advanced toolbar. Press the Ctrl key and click inside any of the selected polygons. The Fill Hole message appears and notifies you that all holes in the selected features will be filled.


1 Answers

Both grd and range are of the class SpatialPolygonsDataFrame.

range has a structure like this:

> str(range)
Formal class 'SpatialPolygonsDataFrame' [package "sp"] with 5 slots
  ..@ data       :'data.frame': 3 obs. of  17 variables:
  .. ..$ OBJECTID  : int [1:3] 725 726 727
  .. ..$ SPCRECID  : int [1:3] 3313 3313 3313
  .. ..$ DATE_     : Factor w/ 1 level "19/10/2009": 1 1 1
  .. ..$ SCINAME   : Factor w/ 1 level "Synthliboramphus antiquus": 1 1 1
  .. ..$ SOURCE    : Factor w/ 1 level "Gaston & Jones, 1998; Brazil, 2009": 1 1 1
  .. ..$ PRESENCE  : int [1:3] 1 1 1
  .. ..$ ORIGIN    : int [1:3] 1 1 1
  .. ..$ SEASONAL  : int [1:3] 2 1 3
  .. ..$ DATA_SENS : Factor w/ 0 levels: NA NA NA
  .. ..$ SENS_COMM : Factor w/ 0 levels: NA NA NA
  .. ..$ COMPILER  : Factor w/ 1 level "Philip Taylor (BirdLife International)": 1 1 1
  .. ..$ TAX_COM   : Factor w/ 0 levels: NA NA NA
  .. ..$ DIST_COM  : Factor w/ 0 levels: NA NA NA
  .. ..$ REVIEWERS : Factor w/ 0 levels: NA NA NA
  .. ..$ CITATION  : Factor w/ 1 level "BirdLife International and Natureserve (2011) Bird species distribution maps of the world.": 1 1 1
  .. ..$ SHAPE_Leng: num [1:3] 190 209 1240
  .. ..$ SHAPE_Area: num [1:3] 39 16.6 600.8
  .. ..- attr(*, "data_types")= chr [1:17] "N" "N" "C" "C" ...
  ..@ polygons   :List of 3
  .. ..$ :Formal class 'Polygons' [package "sp"] with 5 slots
  .. .. .. ..@ Polygons :List of 138
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] 178.6 51.8
  .. .. .. .. .. .. ..@ area   : num 4.58
  .. .. .. .. .. .. ..@ hole   : logi FALSE
  .. .. .. .. .. .. ..@ ringDir: int 1
  .. .. .. .. .. .. ..@ coords : num [1:49, 1:2] 180 180 179 179 179 ...
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] -172.2 52.6

... and so on.

grd has a structure like this:

> str(grd)
Formal class 'SpatialPolygonsDataFrame' [package "sp"] with 5 slots
  ..@ data       :'data.frame': 40680 obs. of  12 variables:
  .. ..$ ID        : int [1:40680] 361 362 363 364 365 366 367 368 369 370 ...
  .. ..$ X_COORD   : num [1:40680] -180 -178 -178 -176 -176 ...
  .. ..$ Y_COORD   : num [1:40680] 79.1 79.1 79.1 79.1 79.1 ...
  .. ..$ ID_1      : Factor w/ 13713 levels "10","10000","10001",..: NA NA NA NA NA NA NA NA NA NA ...
  .. ..$ grdID     : int [1:40680] 1 2 3 4 5 6 7 8 9 10 ...
  .. ..$ ROW       : int [1:40680] 1 1 1 1 1 1 1 1 1 1 ...
  .. ..$ COL       : int [1:40680] 1 2 3 4 5 6 7 8 9 10 ...
  .. ..$ AREA      : num [1:40680] 12364 12364 12364 12364 12364 ...
  .. ..$ PERIMETER : num [1:40680] 445 445 445 445 445 ...
  .. ..$ MAJORITY  : int [1:40680] 0 0 0 0 0 0 0 0 0 0 ...
  .. ..$ PROP0_0062: num [1:40680] 0 0 0 0 0 0 0 0 0 0 ...
  .. ..$ AREA_KM2  : num [1:40680] 0 0 0 0 0 0 0 0 0 0 ...
  .. ..- attr(*, "data_types")= chr [1:12] "N" "N" "N" "C" ...
  ..@ polygons   :List of 40680
  .. ..$ :Formal class 'Polygons' [package "sp"] with 5 slots
  .. .. .. ..@ Polygons :List of 1
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] -179.5 79.1
  .. .. .. .. .. .. ..@ area   : num 5.3
  .. .. .. .. .. .. ..@ hole   : logi FALSE
  .. .. .. .. .. .. ..@ ringDir: int 1
  .. .. .. .. .. .. ..@ coords : num [1:13, 1:2] -179 -180 -180 -180 -180 ...
  .. .. .. ..@ plotOrder: int 1
  .. .. .. ..@ labpt    : num [1:2] -179.5 79.1

... and so on.

I found this help from Roger Bivand and this worked for two out of three problematic files:

>slot(shape.input, "polygons") <- lapply(slot(shape.input, "polygons"), checkPolygonsHoles)
>range.grd.overlay <- grd[!is.na(over(grd, range)), ]@data$grdID # get gridcells of that overlay with range polygon

but for the third file the fixing did not work. R would freeze until I eventually pressed Esc, after which I would get the same error message when trying to make the spatial overlay:

 Error in createPolygonsComment(p) : 
   rgeos_PolyCreateComment: orphaned hole, cannot find containing polygon for hole at index 13
like image 104
user1012744 Avatar answered Oct 07 '22 03:10

user1012744