11

I am in the process of building a ggplot choropleth map of population in administrative areas in Wales. I have downloaded the Boundary-Line data from the Ordnance Survey and extracted what seems to be the right shapefile (community_ward_region.shp). Using R, I have got as far as reading in the shapefile.

require(maptools)
shape <- readShapePoly(wards)
str(shape)

Which gives me this promising output:

Formal class 'SpatialPolygonsDataFrame' [package "sp"] with 5 slots
  ..@ data       :'data.frame': 1690 obs. of  4 variables:
  .. ..$ NAME      : Factor w/ 1507 levels "Abbey Cwmhir",..: 969 90 111 200 441 477 1455 249 255 305 ...
  .. ..$ DESCRIPTIO: Factor w/ 4 levels "COMMUNITY","COMMUNITY WARD",..: 2 2 1 1 2 2 2 2 1 1 ...
  .. ..$ COMMUNITY : Factor w/ 858 levels "Abbey Cwmhir",..: 67 67 81 128 152 152 152 152 157 190 ...
  .. ..$ FILE_NAME : Factor w/ 23 levels "ABERTAWE_-_SWANSEA",..: 1 1 1 1 1 1 1 1 1 1 ...
  .. ..- attr(*, "data_types")= chr [1:4] "C" "C" "C" "C"
  ..@ polygons   :List of 1690
  .. ..$ :Formal class 'Polygons' [package "sp"] with 5 slots
  .. .. .. ..@ Polygons :List of 1
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] 259009 188524
  .. .. .. .. .. .. ..@ area   : num 1923892
  .. .. .. .. .. .. ..@ hole   : logi FALSE
  .. .. .. .. .. .. ..@ ringDir: int 1
  .. .. .. .. .. .. ..@ coords : num [1:1629, 1:2] 259413 259420 259427 259427 259432 ...
  .. .. .. ..@ plotOrder: int 1
  .. .. .. ..@ labpt    : num [1:2] 259009 188524
  .. .. .. ..@ ID       : chr "0"
  .. .. .. ..@ area     : num 1923892

Now if I do this:

bar <- fortify(shape, region = "NAME")

I get a nice data frame called bar that looks pretty much as I expected:

> str(bar)
'data.frame':   4744053 obs. of  7 variables:
 $ long : num  302962 302970 302974 303013 303015 ...
 $ lat  : num  280066 280076 280078 280097 280105 ...
 $ order: int  1 2 3 4 5 6 7 8 9 10 ...
 $ hole : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
 $ piece: Factor w/ 29 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ group: Factor w/ 1762 levels "Abbey Cwmhir.1",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ id   : chr  "Abbey Cwmhir" "Abbey Cwmhir" "Abbey Cwmhir" "Abbey Cwmhir" ..

However, this is a large data frame and ggplot runs out of puff when trying to display it. In reality I only want to look at one area at a time. It looks as if the FILE_NAME factor in the shape object is what I want as it mostly corresponds to counties and the major conurbations.

> unique(shape@data$FILE_NAME)
 [1] ABERTAWE_-_SWANSEA
 [2] BLAENAU_GWENT_-_BLAENAU_GWENT
 [3] BRO_MORGANNWG_-_THE_VALE_OF_GLAMORGAN
 [4] CAERDYDD_-_CARDIFF
 [5] CAERFFILI_-_CAERPHILLY
 [6] CASNEWYDD_-_NEWPORT
 [7] CASTELL-NEDD_PORT_TALBOT_-_NEATH_PORT_TALBOT
 [8] CONWY_-_CONWY
 [9] GWYNEDD_-_GWYNEDD
[10] MERTHYR_TUDFUL_-_MERTHYR_TYDFIL
[11] PEN-Y-BONT_AR_OGWR_-_BRIDGEND
[12] POWYS_-_POWYS
[13] RHONDDA_CYNON_TAF_-_RHONDDA_CYNON_TAFF
[14] SIR BENFRO - PEMBROKESHIRE
[15] SIR_BENFRO_-_PEMBROKESHIRE
[16] SIR_CEREDIGION_-_CEREDIGION
[17] SIR_DDINBYCH_-_DENBIGHSHIRE
[18] SIR_FYNWY_-_MONMOUTHSHIRE
[19] SIR_GAERFYRDDIN_-_CARMARTHENSHIRE
[20] SIR_YNYS_MON_-_ISLE_OF_ANGLESEY
[21] SIR_Y_FFLINT_-_FLINTSHIRE
[22] TOR-FAEN_-_TORFAEN
[23] WRECSAM_-_WREXHAM
23 Levels: ABERTAWE_-_SWANSEA ... WRECSAM_-_WREXHAM

Q. How can I select only a subset of the data from the shape object I extract from the shapefile? For example, only the POWYS_-_POWYS parts? If I can somehow include the FILE_NAME in the data frame that is created with fortify then I could easily subset the bar data frame but I don't know how to do that. Or is there a way to use fortify to extract only parts of the object?

1 Answer 1

5

Select a subset of the shapefile data using indexing:

sub.shape <- shape[shape$FILE_NAME == "POWYS_-_POWYS",]

fortify(sub.shape) will then give you a much reduced dataframe.

2
  • Thank you, checking now, looks promising. Should that perhaps be a ",]" at the end there? Oct 20, 2012 at 13:42
  • 1
    @slowlearner - Whoops, yes, missing comma, my perpetual error ...
    – Simbamangu
    Oct 20, 2012 at 19:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.