How can I make the output from tapply() into a data.frame

I have a data.frame in R that looks like this:

      score    rms  template   aln_id       description
1  -261.410  4.951 2f22A.pdb  2F22A_1 S_00001_0000002_0
2  -231.987 21.813 1wb9A.pdb  1WB9A_开发者_C百科4 S_00002_0000002_0
3  -263.722  4.903 2f22A.pdb  2F22A_3 S_00003_0000002_0
4  -269.681 17.732 1wbbA.pdb  1WBBA_6 S_00004_0000002_0
5  -258.621 19.098 1rxqA.pdb  1RXQA_3 S_00005_0000002_0
6  -246.805  6.889 1rxqA.pdb 1RXQA_15 S_00006_0000002_0
7  -281.300 16.262 1wbdA.pdb 1WBDA_11 S_00007_0000002_0
8  -271.666  4.193 2f22A.pdb  2F22A_2 S_00008_0000002_0
9  -277.964 13.066 1wb9A.pdb  1WB9A_5 S_00009_0000002_0
10 -261.024 17.153 1yy9A.pdb  1YY9A_2 S_00001_0000003_0

I can calculate summary statistics on the data.frame like this:

> tapply( d$score, d$template, mean )
1rxqA.pdb 1wb9A.pdb 1wbbA.pdb 1wbdA.pdb 1yy9A.pdb 2f22A.pdb 
-252.7130 -254.9755 -269.6810 -281.3000 -261.0240 -265.5993 

Is there an easy way that I coerce this output back into a data.frame? I'd like for it to have these two columns:

d$template
mean

I love tapply, but right now I'm cutting and pasting the results from tapply into a text file and hacking it up a bit to get the summary statistics that I want with appropriate names. This feels very wrong, and I'd like to do something better!

---

There are a lot of different ways to transform the output from a tapply call into a data.frame.

But it's much simpler to avoid the call to tapply in the first place and substitute that with a call to a similar function that returns a data frame instead of a vector:

more specifically:

• tapply returns a vector

• aggregate returns a data frame

so just change your function call from tapply to aggregate, like so:

data(iris)     # in 'datasets' just call 'data' and pass in 'iris' as an argument

tx = tapply(iris$Sepal.Length, list(iris$Species), mean)
# returns: versicolor  virginica 
             5.94       6.59 

class(tx)
# returns: vector

tx = aggregate(iris$Sepal.length, list(iris$Species), mean)
# returns:
         Group.1    x
     1 versicolor 5.94
     2  virginica 6.59


class(tx)
# returns: data.frame

---

You can try this:

mn <- tapply(d$score,d$template,mean)
df <- data.frame(template=names(mn),mean=mn)

---

library(plyr)
ddply(d, "template", summarise, mean = mean(score))