creating table of averages

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20080909/88172bc0/attachment.pl>
Maybe something like this:

by(df[,c(77,81,86,90,94,98,101,106)],df$category,apply,2,mean)

...which would then need to be reformatted into a data frame (there is
probably an easy way to do this which I don't know).

aggregate seems like a more reasonable choice, but the function for
aggregate must return scalars, not rows...tapply doesn't take data.frame
inputs. Maybe someone else has a suggestion?

--Adam

Dear Colleagues,

I have a dataframe with variables:

 [1] "ID"             "category"       "a11"            "a12"
"a13"            "a21"
 [7] "a22"            "a23"            "a31"            "a32"
"b11"            "b12"
[13] "b13"            "b21"            "b31"            "b32"
"b33"            "b41"
[19] "b42"            "c11"            "c12"            "c21"
"c22"            "c23"
[25] "c31"            "c32"            "c33"            "d11"
"d12"            "d13"
[31] "d14"            "d21"            "d22"            "d23"
"d24"            "d25"
[37] "d31"            "d32"            "d33"            "e11"
"e12"            "e13"
[43] "e21"            "e22"            "e23"            "e31"
"e32"            "e33"
[49] "f11"            "f12"            "f13"            "f14"
"f21"            "f22"
[55] "f23"            "f24"            "g11"            "g12"
"g13"            "g14"
[61] "g21"            "g22"            "g23"            "g24"
"g31"            "g32"
[67] "g33"            "g41"            "g42"            "g43"
"h11"            "h12"
[73] "h13"            "h21"            "h22"            "h23"
"C1.Employ"      "SC11.Ops"
[79] "SC12.Unit"      "SC13.Nonadvers" "C2.Enterprise"  "SC21.Structure"
"SC22.Gov"       "SC23.Culture"
[85] "SC24.Stratcomm" "C3.Manage"      "SC31.Resource"  "SC32.Change"
"SC33.Continue"  "C4.Stratthink"
[91] "SC41.Vision"    "SC42.Decision"  "SC43.Adapt"     "C5.Lead"
"SC51.Develop"   "SC52.Care"
[97] "SC53.Diversity" "C6.Foster"      "SC61.Teams"     "SC62.Negotiate"
"C7.Embody"      "SC71.Ethical"
[103] "SC72.Follower"  "SC73.Warrior"   "SC74.Develop"   "C8.Comm"
"C81.Speak"      "C82.Listen"
[109] "OverallImp"

The variable "category" has four values: Regular, CCM, CFM, and Other

I'd like to create a table like this to feed into barplot2:

row.name  C1.Employ C2.Enterprise  C3.Manage  C4.Stratthink  C5.Lead
C6.Foster  C7.Embody  C8.Comm
Regular         3.68              4.27             3.22
etc......
CCM             4.32              4.56              etc.....
CFM              etc.........
Other             etc.........

So far, I have been able to get this far:

mean(subset(impchiefs08,category=="Regular",select=c(C1.Employ,C2.Enterprise,C3.Manage,C4.Stratthink,C5.Lead,C6.Foster,C7.Embody,C8.Comm
)))
   C1.Employ C2.Enterprise     C3.Manage C4.Stratthink       C5.Lead
C6.Foster     C7.Embody       C8.Comm
    3.600000      3.851111      4.482222      4.346667      4.608889
4.444444      4.602222      4.493333

But I am stumped as to how to get what I want.

Thanks in advance.

Larry

	[[alternative HTML version deleted]]

______________________________________________
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.

Maybe something like this:

by(df[,c(77,81,86,90,94,98,101,106)],df$category,apply,2,mean)

...which would then need to be reformatted into a data frame (there is
probably an easy way to do this which I don't know).
sparseby() in the reshape package is more flexible than by(). If the 
function returns a vector with a consistent length, you'll get a 
dataframe with columns corresponding to its entries.

Duncan Murdoch
aggregate seems like a more reasonable choice, but the function for
aggregate must return scalars, not rows...tapply doesn't take data.frame
inputs. Maybe someone else has a suggestion?

--Adam

On Tue, 9 Sep 2008, Lawrence Hanser wrote: