On R 2.3.1 I was surprised to see the following results from
summary.default():
R> x <- scan()
1: 12148 10426 10912 9116 13226 11663 11781 10680 8457 10788 12605
10591 11040
14: 10815 12962 11644 10047 10478 10108 12353 11778 11092 11673 8758
11145 11495
27:
Read 26 items
R> summary(x)
Min. 1st Qu. Median Mean 3rd Qu. Max.
8460 10500 11100 11100 11800 13200
R> max(x)
[1] 13226
Clearly, these are being rounded; help(summary) shows the digits=
option. If I now ask for all five digits (n.b. I have to know the
magnitude of the numbers to know how many integer digits) I get:
R> summary(x, digits=5)
Min. 1st Qu. Median Mean 3rd Qu. Max.
8457 10506 11066 11068 11752 13226
which now shows the actual maximum etc.
help(summary) shows options: digits = max(3, getOption("digits")-3)
R> getOption("digits")
[1] 5
So in my case I get three digits... but I had thought this applied to
decimal digits, not integer digits, and that rounding would only
occur in decimal places, according to the digits option.
Question: is this behaviour of summary.default what is expected? If
so, it seems non-intuitive to me.
Thank you,
D. G. Rossiter
Senior University Lecturer
Department of Earth Systems Analysis
International Institute for Geo-Information Science and Earth
Observation (ITC)
Enschede, The Netherlands