[Rd] Please make Pre-3.1 read.csv (type.convert) behavior available

[knormoyle]

Hi Duncan,
I'm with Tom, don't want to be redundant but here's some extra info.
This made me think that the problem is not a 'theshold'. Any thoughts.
Also, if the "bad" number strings are entered at the R command prompt, they
are parsed correctly as the expected number. (not factors)

thanks,
-kevin


this works

0x1.ffadp-1

> df = read.csv("bad1.csv", header=F)
> str(df)
'data.frame':    1 obs. of  1 variable:
 $ V1: num 0.999


but this doesn't

0x1.ffa000000000dp-1

> df = read.csv("bad1.csv", header=F)
> str(df)
'data.frame':    1 obs. of  1 variable:
 $ V1: Factor w/ 1 level "0x1.ffa000000000dp-1 ": 1


this also works, which is one less trailing zero.

0x1.ffa00000000dp-1

>  df = read.csv("bad1.csv", header=F)
> str(df)
'data.frame':    1 obs. of  1 variable:
 $ V1: num 0.999 



--
View this message in context: http://r.789695.n4.nabble.com/Please-make-Pre-3-1-read-csv-type-convert-behavior-available-tp4689507p4689553.html
Sent from the R devel mailing list archive at Nabble.com.