R help - Aug 2012 - Using unicode symbol has unexpected results in levels of factor object

Dear all,

When I use a unicode symbol in the labels for a factor object, the corresponding
level does not display as expected. However, using levels() on the factor
returns the desired output. I noticed the discrepancy when the legend labels
from a call to ggplot() did not display the desired symbol, but an explicitly
built legend using the same labels did.

Example (I am trying to get the less than or equal to symbol): 
> .df <- data.frame(afp = c(0,0,1,1), time=c(0,2,0,1), surv=c(1, 0.5, 1,
0.4))
> afpLabels <- c("AFP \u2264 16", "AFP > 16")
> afpStrata <- factor(.df$afp, labels=afpLabels)
> afpStrata
[1] AFP ? 16 AFP ? 16 AFP > 16 AFP > 16
Levels: AFP = 16 AFP > 16

The first level is reported as "AFP = 16".
> levels(afpStrata)
[1] "AFP ? 16" "AFP > 16"
> 
The desired result is produced with levels().


The code below shows this issue in context through calls to ggplot() if you
don't mind loading all the libraries.
> library(ggplot2)
> library(gridExtra)
> library(plyr)
> 
> ggplot(.df, aes(time, surv)) + geom_step(aes(color = afpStrata), size =
1.0)
> 
> ggplot(.df, aes(time, surv)) + geom_step(aes(color = afpStrata), size =
1.0)  +
+ scale_colour_hue(breaks=afpLabels, labels=afpLabels)
> 
I am running a pre-compiled version of R on Windows 7
(64-bit).
> sessionInfo()
R version 2.15.1 (2012-06-22)
Platform: x86_64-pc-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  methods  
[8] base     

other attached packages:
[1] plyr_1.7.1    gridExtra_0.9 ggplot2_0.9.1

loaded via a namespace (and not attached):
 [1] colorspace_1.1-1   dichromat_1.2-4    digest_0.5.2       labeling_0.1      
 [5] MASS_7.3-18        memoise_0.1        munsell_0.3        proto_0.3-9.2     
 [9] RColorBrewer_1.0-5 reshape2_1.2.1     scales_0.2.1       stringr_0.6.1     
[13] tools_2.15.1      
> 
Sincerely,
Kristin Berry

	[[alternative HTML version deleted]]

On Aug 9, 2012, at 06:53 , Wyatt, Kristin M wrote:
> Dear all,
> 
> When I use a unicode symbol in the labels for a factor object, the
corresponding level does not display as expected. However, using levels() on the
factor returns the desired output. I noticed the discrepancy when the legend
labels from a call to ggplot() did not display the desired symbol, but an
explicitly built legend using the same labels did.
> 
> Example (I am trying to get the less than or equal to symbol): 
> 
>> .df <- data.frame(afp = c(0,0,1,1), time=c(0,2,0,1), surv=c(1, 0.5,
1, 0.4))
>> afpLabels <- c("AFP \u2264 16", "AFP > 16")
>> afpStrata <- factor(.df$afp, labels=afpLabels)
>> afpStrata
> [1] AFP ? 16 AFP ? 16 AFP > 16 AFP > 16
> Levels: AFP = 16 AFP > 16
> 
> The first level is reported as "AFP = 16".
> 
>> levels(afpStrata)
> [1] "AFP ? 16" "AFP > 16"
>> 
> 
> The desired result is produced with levels().
> 
> 
> The code below shows this issue in context through calls to ggplot() if you
don't mind loading all the libraries.
> 
>> library(ggplot2)
>> library(gridExtra)
>> library(plyr)
>> 
>> ggplot(.df, aes(time, surv)) + geom_step(aes(color = afpStrata), size =
1.0)
>> 
>> ggplot(.df, aes(time, surv)) + geom_step(aes(color = afpStrata), size =
1.0)  +
> + scale_colour_hue(breaks=afpLabels, labels=afpLabels)
>> 
> 
> I am running a pre-compiled version of R on Windows 7 (64-bit).
>> sessionInfo()
> R version 2.15.1 (2012-06-22)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
For whatever it is worth, this works fine (both examples) under OSX Snow
Leopard.

Looking at the code for print.factor, I would strongly suspect that the culprit
is the line

        n <- length(lev <- encodeString(levels(x), quote = ifelse(quote, 
            "\"", "")))

which figures since you are in a .1252 locale, not .utf8 (or UTF-8 or ...). 

Over to the Windows/locale/charset experts...

-- 
Peter Dalgaard, Professor
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com