R help - Mar 2008 - summarizing replicates with multiple treatments

I have a dataframe with several different treatment variables, and
would like to calculate the mean and standard deviation of the
replicates for each day and treatment variable.  It seems like it
should be easy, but I've only managed to do it for one treatment at a
time using subset and tapply.  Here is an example dataset:
> `exampledata` <-
structure(list(day = c(1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L,
2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L), treat = structure(c(1L, 1L,
1L, 2L, 2L, 2L, 1L, 1L, 1L, 2L, 2L, 2L, 1L, 1L, 1L, 2L, 2L, 2L
), .Label = c("a", "b"), class = "factor"),
replicate = c(1L,
2L, 3L, 1L, 2L, 3L, 1L, 2L, 3L, 1L, 2L, 3L, 1L, 2L, 3L, 1L, 2L,
3L), height = c(1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 2.1, 2.2, 2.3,
2.4, 2.5, 2.6, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6), weight = c(11.1,
11.2, 11.3, 11.4, 11.5, 11.6, 12.1, 12.2, 12.3, 12.4, 12.5, 12.6,
13.1, 13.2, 13.3, 13.4, 13.5, 13.6)), .Names = c("day",
"treat",
"replicate", "height", "weight"), class =
"data.frame", row.names = c(NA,
-18L))
> exampledata
   day treat replicate height weight
1    1     a         1    1.1   11.1
2    1     a         2    1.2   11.2
3    1     a         3    1.3   11.3
4    1     b         1    1.4   11.4
5    1     b         2    1.5   11.5
6    1     b         3    1.6   11.6
7    2     a         1    2.1   12.1
8    2     a         2    2.2   12.2
9    2     a         3    2.3   12.3
10   2     b         1    2.4   12.4
11   2     b         2    2.5   12.5
12   2     b         3    2.6   12.6
13   3     a         1    3.1   13.1
14   3     a         2    3.2   13.2
15   3     a         3    3.3   13.3
16   3     b         1    3.4   13.4
17   3     b         2    3.5   13.5
18   3     b         3    3.6   13.6

I would like to combine the replicates and get a dataframe like:

day	treat	height.mean	height.sd	weight.mean	weight.sd
1	a	1.2	0.1	11.2	0.1
1	b	1.5	0.1	11.5	0.1
2	a	2.2	0.1	12.2	0.1
2	b	2.5	0.1	12.5	0.1
3	a	3.2	0.1	13.2	0.1
3	b	3.5	0.1	13.5	0.1

or two dataframes, one with means and the other with standard deviations.

Thus far I have been doing it a piece at a time, like below (extra
verbose since tapply doesn't accept the data= argument!), but would
like to do it for all the measurement columns and all the treatments
in one go.  Thanks!
>
tapply(exampledata[exampledata$treat=="a",]$height,exampledata[exampledata$treat=="a",]$day,mean)
  1   2   3
1.2 2.2 3.2
>

Try this:

library(doBy)
summaryBy(. ~ day + treat, exampledata, FUN = c(mean, sd))


On Tue, Mar 4, 2008 at 5:24 PM, Levi Waldron <leviwaldron at gmail.com>
wrote:
> I have a dataframe with several different treatment variables, and
> would like to calculate the mean and standard deviation of the
> replicates for each day and treatment variable.  It seems like it
> should be easy, but I've only managed to do it for one treatment at a
> time using subset and tapply.  Here is an example dataset:
>
> > `exampledata` <-
> structure(list(day = c(1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L,
> 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L), treat = structure(c(1L, 1L,
> 1L, 2L, 2L, 2L, 1L, 1L, 1L, 2L, 2L, 2L, 1L, 1L, 1L, 2L, 2L, 2L
> ), .Label = c("a", "b"), class = "factor"),
replicate = c(1L,
> 2L, 3L, 1L, 2L, 3L, 1L, 2L, 3L, 1L, 2L, 3L, 1L, 2L, 3L, 1L, 2L,
> 3L), height = c(1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 2.1, 2.2, 2.3,
> 2.4, 2.5, 2.6, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6), weight = c(11.1,
> 11.2, 11.3, 11.4, 11.5, 11.6, 12.1, 12.2, 12.3, 12.4, 12.5, 12.6,
> 13.1, 13.2, 13.3, 13.4, 13.5, 13.6)), .Names = c("day",
"treat",
> "replicate", "height", "weight"), class =
"data.frame", row.names = c(NA,
> -18L))
>
> > exampledata
>   day treat replicate height weight
> 1    1     a         1    1.1   11.1
> 2    1     a         2    1.2   11.2
> 3    1     a         3    1.3   11.3
> 4    1     b         1    1.4   11.4
> 5    1     b         2    1.5   11.5
> 6    1     b         3    1.6   11.6
> 7    2     a         1    2.1   12.1
> 8    2     a         2    2.2   12.2
> 9    2     a         3    2.3   12.3
> 10   2     b         1    2.4   12.4
> 11   2     b         2    2.5   12.5
> 12   2     b         3    2.6   12.6
> 13   3     a         1    3.1   13.1
> 14   3     a         2    3.2   13.2
> 15   3     a         3    3.3   13.3
> 16   3     b         1    3.4   13.4
> 17   3     b         2    3.5   13.5
> 18   3     b         3    3.6   13.6
>
> I would like to combine the replicates and get a dataframe like:
>
> day     treat   height.mean     height.sd       weight.mean     weight.sd
> 1       a       1.2     0.1     11.2    0.1
> 1       b       1.5     0.1     11.5    0.1
> 2       a       2.2     0.1     12.2    0.1
> 2       b       2.5     0.1     12.5    0.1
> 3       a       3.2     0.1     13.2    0.1
> 3       b       3.5     0.1     13.5    0.1
>
> or two dataframes, one with means and the other with standard deviations.
>
> Thus far I have been doing it a piece at a time, like below (extra
> verbose since tapply doesn't accept the data= argument!), but would
> like to do it for all the measurement columns and all the treatments
> in one go.  Thanks!
>
> >
tapply(exampledata[exampledata$treat=="a",]$height,exampledata[exampledata$treat=="a",]$day,mean)
>  1   2   3
> 1.2 2.2 3.2
> >
>
> ______________________________________________
> R-help at 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.
>

On Tue, Mar 4, 2008 at 5:32 PM, Gabor Grothendieck
<ggrothendieck at gmail.com> wrote:
> Try this:
>
>  library(doBy)
>  summaryBy(. ~ day + treat, exampledata, FUN = c(mean, sd))
Outstanding, so much better.  Thanks.