R devel - Jul 2001 - ordered factors in tree package - bug? (PR#1025)

I am new to R, and didn't know which list to send this to, since it is a
bug report about a package, not about core R...

I have created a regression tree using 4 predictors: 3 are unordered
(binary) predictors, and the last is a date (integer), which I declare
to be an ordered factor. However, the tree treats the date as if it were
un-ordered, splitting into non-consecutive subsets.

My code

dat <- read.table("/home/cs/murphyk/R/Eugene/102.dat", header=TRUE)
dat$machine <- factor(dat$machine)
dat$TIM <- factor(dat$TIM)
dat$lid <- factor(dat$lid)
#dat$date <- factor(dat$date, ordered=TRUE)
dat$date <- ordered(dat$date)
tr <- tree(TRES ~ ., dat)

produces

 1) root 100 0.149300 0.2345  
   2) TIM: 111 47 0.020650 0.2011  
     4) date: 9,11 11 0.002673 0.1845 *
     5) date: 6,7,8,10,13,14,15,16,17,18,19 36 0.014060 0.2061 *
   3) TIM: 222 53 0.029490 0.2642  
     6) date: 6,7,11,12,14,15,18,19 28 0.010670 0.2539  
      12) lid: A 6 0.001350 0.2350 *
      13) lid: B 22 0.006582 0.2591 *
     7) date: 8,9,10,13,16,17 25 0.012620 0.2756  
      14) machine: 101 14 0.005750 0.2850 *
      15) machine: 102 11 0.004055 0.2636 *

and yet
> dat$date
  [1] 6  6  6  6  6  7  7  7  7  7  7  7  8  8  8  9  9  9  9  9  9  9 
9  9  10
 [26] 10 10 10 10 10 11 11 11 11 11 11 11 11 12 12 12 13 13 13 13 13 13
13 13 13
 [51] 13 13 14 14 14 14 14 14 14 15 15 15 15 15 15 15 15 15 15 16 16 16
16 16 16
 [76] 16 16 17 17 17 17 17 17 17 17 17 17 17 17 17 18 18 18 18 18 19 19
19 19 19
Levels:  6 < 7 < 8 < 9 < 10 < 11 < 12 < 13 < 14 < 15
< 16 < 17 < 18 < 19 
> is.ordered(dat$date)
[1] TRUE



Also, how do I deal with dates of the form dd/mm/yy, instead of just
integers? (In the above file, I used perl to extract the day, since I
new month and year were constant.)

I have attached the text file 102.dat, so you can easily reproduce the
above bug. I am using R 1.2.3 on linux. For future reference, is it
considered bad form to send attachments?


Kevin