Skip to content

colorkey print order is incorrect??

3 messages · Jim Burke, Roger Bivand, Edzer Pebesma

Original
Jim Burke
# 
Question. When I use the code below to print
a list of values to the right of my graph,
the colorkey list appears unusual.  At the top
the list of percentages tops out at 90.87%
instead of the expected 100%. The 100% appears
at the bottom below the lowest percent 14.29%.

The numbers of tx3_sp CK_PCT are read in as follows.
ID is the PCT and CK_PCT is what's plotted.
 > tx3_sp
An object of class "SpatialPolygonsDataFrame"
Slot "data":
    PCT    CD   COMM   JP     LD    SB    SS CK_TOT  CK_PCT TOT_VOTE
1  1120 CD-03 COMM01 JP-3 LD-102 SB-12 SS-16   1693  90.87%     1863
2  1121 CD-05 COMM01 JP-3 LD-102 SB-12 SS-16    634  54.28%     1168
3  1124 CD-05 COMM01 JP-3 LD-102 SB-12 SS-16    584  43.39%     1346
4  1135 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    815  33.51%     2432
5  1136 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    446  52.84%      844
6  1719 CD-03 COMM01 JP-3 LD-102 SB-12 SS-16   1392  68.17%     2042
7  1720 CD-03 COMM01 JP-3 LD-102 SB-12 SS-16    523  66.20%      790
8  1723 CD-03 COMM01 JP-3 LD-102 SB-12 SS-16    788  82.34%      957
9  1724 CD-03 COMM01 JP-3 LD-102 SB-12 SS-16    994  69.41%     1432
10 1725 CD-32 COMM01 JP-3 LD-102 SB-12 SS-16     31  59.62%       52
11 1802 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    999  44.20%     2260
12 1803 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    672  44.24%     1519
13 1804 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    280  39.94%      701
14 1805 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    654  48.12%     1359
15 1806 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    495  37.61%     1316
16 1807 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    695  40.81%     1703
17 1809 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    563  47.11%     1195
18 1810 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    461  42.92%     1074
19 1811 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    779  42.61%     1828
20 1812 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    585  37.77%     1549
21 1816 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    311  32.16%      967
22 1817 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    528  44.18%     1195
23 1819 CD-32 COMM01 JP-3 LD-102 SB-12 SS-08    338  47.88%      706
24 1820 CD-32 COMM01 JP-3 LD-102 SB-12 SS-16    305  70.11%      435
25 1821 CD-03 COMM01 JP-3 LD-102 SB-12 SS-16    617  74.07%      833
26 2115 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16      1  14.29%        7
27 2121 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16    646  50.91%     1269
28 2122 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16    120  60.00%      200
29 2126 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16    511  57.55%      888
30 2127 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16    522  41.04%     1272
31 2128 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16    390  49.49%      788
32 2129 CD-05 COMM02 JP-2 LD-102 SB-12 SS-16    243  41.33%      588
33 2150 CD-03 COMM02 JP-2 LD-102 SB-12 SS-16      2 100.00%        2
34 2151 CD-05 COMM02 JP-2 LD-102 SB-12 SS-16     36  38.71%       93
35 2200 CD-03 COMM02 JP-3 LD-102 SB-12 SS-16    959  88.39%     1085
36 2204 CD-05 COMM02 JP-3 LD-102 SB-12 SS-16    363  90.07%      403
37 2205 CD-03 COMM02 JP-3 LD-102 SB-12 SS-16    921  90.38%     1019
38 2206 CD-05 COMM02 JP-3 LD-102 SB-12 SS-16    704  45.77%     1538



Is there any way to move the 100% to the top
when the colorkey legend prints?

    spplot(tx3_sp, c("CK_PCT"),
        names.attr = c("Carol Kent 2008"),
        colorkey=list(height=1.0, width=0.6),
        main = "HD102 - Carol Kent",
        sp.layout = list("sp.text",coordinates(tx3_sp),
        as.character(tx3_sp$PCT,cex = .02),
        as.table = TRUE))

See the attached jpg file. I hope this goes through.

Forgive my dumb question as I am new to R. I like
R after reading the NYTIMES article last week;
"Data Analysts Mesermized by Power of R". Query
the NYTIMES for this. And read the "Are U Ready
for R" article with its comments as well.

Thanks,
Jim Burke



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20090119/1307384b/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PercentOrder.jpg
Type: image/jpeg
Size: 55874 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20090119/1307384b/attachment.jpg>
Roger Bivand
#
On Mon, 19 Jan 2009, Jim Burke wrote:

            

      
      
      
    

        
Yes, very nice! Your variable CK_PCT is clearly a categorical variable, 
with a % in it, and has been sorted correctly, as "100.00%" < "14.29%" in 
alphanumeric sort order. Maybe:

tx3_sp$nCK_PCT <- (100*tx3_sp$CK_TOT)/tx3_sp$TOT_VOTE

then map "nCK_PCT" (assuming that CK_TOT and TOT_VOTE are numeric - some 
software makes all fields in say shapefiles into strings)? If they aren't 
numeric (see below), coerce them using as.numeric() first.

You can check the way in which the data are represented by using the str() 
command, here

str(as(tx3_sp, "data.frame"))

will show the representation of your variables (str(tx3_sp) would show the 
geometries too, which you don't need here).

Roger

            

      
      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Edzer Pebesma
    

    

      
      #
    

  


  

      
Jim,

your percentage variable is read as a factor; just check:

 > class(tx3_sp$CK_PCT)

this is because R does not understand % symbols (unless you tell it). 
Convert it to numeric by

 > tx3_sp$CK_PCTN = as.numeric(unlist(strsplit(tx3_sp$CK_PCT,"%")))

and use spplot on the new variable. The as.table=TRUE argument does not 
have an effect when plotting a single variable only.

Thanks; the list is aware of the NY Times article:
https://stat.ethz.ch/pipermail/r-sig-geo/2009-January/004789.html
--
Edzer

        
Jim Burke wrote: