Given the data:
dat1<-read.table(header=TRUE,text="
ID V1 V2 V3 V4
1 6 5 3 2
2 3 2 2 1
3 6 5 3 2
4 12 15 3 2
5 6 8 3 2
6 3 2 4 1
7 6 5 3 3
8 12 15 3 1
9 6 5 3 3
10 3 2 7 5
11 6 5 8 2
12 12 19 3 2
13 6 5 3 2
14 3 4 2 1
15 6 5 6 2
16 12 15 5 2
17 6 5 5 2
18 3 2 8 1
19 6 5 3 9
20 12 15 3 10
21 6 5 3 2
22 3 2 2 11
23 6 5 3 4
24 12 15 9 2
25 6 5 3 2
26 3 2 2 1
27 6 5 3 2
28 12 15 3 2
29 6 8 3 2
30 3 2 4 1
31 6 5 3 3
32 12 15 3 1
33 6 5 3 3
34 3 2 7 5
35 6 5 8 2
36 12 19 3 2
37 6 5 3 2
38 3 4 2 1"))
The following seems to be rather quick (0.394 ms per group of up to 12 rows
in 1201 rows):
do.call(rbind,by(dat1,floor((dat1$ID-1)/12),colMeans))
ID V1 V2 V3 V4
0 6.0 6.272727 6.272727 3.818182 2.181818
1 17.5 6.750000 7.250000 3.833333 4.000000
2 29.5 6.750000 7.000000 4.250000 2.166667
3 37.0 7.000000 9.333333 2.666667 1.666667
If you know the number of rows is a multiple of 12:
dat2<-read.table(header=TRUE,text="
ID V1 V2 V3 V4
1 6 5 3 2
2 3 2 2 1
3 6 5 3 2
4 12 15 3 2
5 6 8 3 2
6 3 2 4 1
7 6 5 3 3
8 12 15 3 1
9 6 5 3 3
10 3 2 7 5
11 6 5 8 2
12 12 19 3 2
13 6 5 3 2
14 3 4 2 1
15 6 5 6 2
16 12 15 5 2
17 6 5 5 2
18 3 2 8 1
19 6 5 3 9
20 12 15 3 10
21 6 5 3 2
22 3 2 2 11
23 6 5 3 4
24 12 15 9 2
25 6 5 3 2
26 3 2 2 1
27 6 5 3 2
28 12 15 3 2
29 6 8 3 2
30 3 2 4 1
31 6 5 3 3
32 12 15 3 1
33 6 5 3 3
34 3 2 7 5
35 6 5 8 2
36 12 19 3 2"))
This is marginally better (0.378 ms per group of 12 rows in 1212 rows, or
4% less time):
do.call(rbind,by(dat2,rep(1:(NROW(dat2)/12),each=12),colMeans))
ID V1 V2 V3 V4
1 6.5 6.75 7.333333 3.750000 2.166667
2 18.5 6.75 6.916667 4.333333 4.000000
3 30.5 6.75 7.333333 3.750000 2.166667
Best,
Mario
On Sat, Jan 5, 2013 at 8:33 AM, Simonas Kecorius <simolas2...@gmail.com>wrote:
> Dear R users, I came up to a problem by taking means (or other summary
> statistics) of a big dataframe.
>
> Suppose we do have a dataframe:
>
> ID V1 V2 V3 V4 ........................ V71
> 1 6 5 3 2 ........................ 3
> 2 3 2 2 1 ........................ 1
> 3 6 5 3 2 ........................ 3
> 4 12 15 3 2 ........................ 100
> ........................................................
> ........................................................
> 288 10 20 30 30 .......................... 499
>
> I need to find out the way, how to calculate a mean of every 12 lines to
> get:
>
> V1 V2 V3 V4
> ........................... V71
> mean from 1 to 7 same as V1 same as V1
> mean from 8 to 14 same as V1 same as V1
> etc.
>
> I can do it column by column using:
>
> y.ts <- ts(y$V1, frequency=12)
> aggregate(y.ts, FUN=mean)
>
> Bu this is a hardcore... Can anyone suggest a better way to compute all the
> dataframe at once and get a result as matrix?
>
> Thank you in advance!
>
> --
> Simonas Kecorius
> **
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
