R's XML example not working - xmlToDataFrame

With the next code:

library(XML)

f = system.file("exampleData", "size.xml", packag开发者_高级运维e = "XML")
doc = xmlParse(f)
z = xmlToDataFrame(f, colClasses = list("integer", "integer", "numeric"))
y = xmlToDataFrame(nodes = xmlChildren(xmlRoot(doc)[["size"]]))

browser()

I get these results :

Browse[1]> z
  age sex number
1   0   0    500
2   0   1    300
3   1   0    200
4   1   1    400
5  10   0     NA
Browse[1]> y
  text
1    0
2    0
3  500

This is from package help and following the manual, what can be wrong? The results for y are not the same for z, why?

Thanks

You didn't copy the entire example. I get the same error when I just copy those two lines from the example, but no error when I copy the entire example from ?xmlToDataFrame .

(I have done this twice and I have not figured out which of the arguments is not being passed properly, and so my working hypothesis is that some connections is not established properly.)

EDIT; (now an entirely different question) Because the root is different than the whole file:

> xmlRoot(doc)[["size"]]
<size>
  <age>0</age>
  <sex>0</sex>
  <number>500</number>
</size> 

EDIT2: You do not need to specify the colClasses:

>  xmlToDataFrame(f)
  age sex number
1   0   0    500
2   0   1    300
3   1   0    200
4   1   1    400
5  10   0   <NA>

And if you want the read.table behavior of guessing at the class from the first five lines of data then you can do a write/read.table operation:

> write.table(xmlToDataFrame(f), file="test.txt", quote=FALSE)
> read.table(file="test.txt")
> str(read.table(file="test.txt"))
'data.frame':   5 obs. of  3 variables:
 $ age   : int  0 0 1 1 10
 $ sex   : int  0 1 0 1 0
 $ number: int  500 300 200 400 NA