[R] summary with variance / sd

David Winsemius dwinsemius at comcast.net
Mon Jan 12 00:24:03 CET 2009


On Jan 11, 2009, at 4:38 PM, Jörg Groß wrote:

> Hi,
>
> I have a data frame and would like to have summary statistics for  
> grouped data.
>
> With summary() I get the central tendencies for the overall data.
>
>
> How can I get descriptive statistics with variances and standard  
> deviations?

In the future, you really should do your own searching before posting  
this sort of basic question.


help.search("standard deviation")  produced reference to a function  
with the obvious name "sd"
Using the same strategy for variance produces a longer list but "var"  
is among them.

> for example my data.frame:
>
> group	x	y
> exp		2	4
> exp		3	5
> exp		2	4
> control	1	2
> control	2	3
> control	1	2

#Create dataframe ( and please note that you are asked to offer  
examples in a form that does not require responders to create the  
objects for you):
  df1 <- read.table(stdin(), header=TRUE)

#Paste in the data:
0: group    x    y
1: exp        2    4
2: exp        3    5
3: exp        2    4
4: control    1    2
5: control    2    3
6: control    1    2
# empty line stops input.

by(data=df1, df1$group, summary)
by(data=df1, df1$group, sd)
by(data=df1, df1$group, var)

Or.... use negative indexing to exclude the first column,  and add  
some annotation to do it in one step

by(data=df1[-1], df1$group, function(x){ list(summary(x), "Group S.D.s  
are ...", sd(x), "Group Variances are ...", var(x) )} )


You could also look at how the pro's do constructed summary()  by  
reviewing the code of:

base:::summary.default
base:::summary.data.frame

-- 
David Winsemius


>
>
>
> now I want tables with summary statistics (variances included) for  
> each group.
>
> Is there an easy way to get this?
>
> ______________________________________________
> 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.




More information about the R-help mailing list