This should work on your data using the 'reshape' package;
> nmr_19_model_aa<-read.table(url("
http://omega.psi.iastate.edu/data/nmr_19_model_aa.txt";),sep<http://omega.psi.iastate.edu/data/nmr_19_model_aa.txt%22),sep>
="")
> name.col<-c("pr","mod","mod_No","aa","aa_po","omega")
> colnames(nmr_19_model_aa)<-name.col
> attach(nmr_19_model_aa)
> nmr_19_model_aa.2<-data.frame
(pr=pr,model=paste(mod,mod_No,sep="_"),aa=paste(aa_po,aa,sep="_"),omega=omega)
> attach(nmr_19_model_aa.2)
> temp<-split(nmr_19_model_aa.2,nmr_19_model_aa.2$pr)
> class(temp)
[1] "list"
> names(temp)
[1] "1A24" "1A57" "1A5J" "1A6X" "1AB7" "1AF8" "1AFI" "1AGG" "1AH9" "1AHL"
"1AJ3" "1AJW" "1AK7" "1ALG" "1AOY" "1APF" "1AUZ"
[18] "1AZJ" "1AZK"
> # show how to do it with one
> require(reshape)
[1] TRUE
> x.ans <- lapply(temp, function(.df){
+ x.r <- melt(.df) # setup for 'casting'
+ x.c <- as.matrix(cast(x.r, aa ~ model)) # make a matrix
+ cbind(x.c, mean=apply(x.c, 1, mean), sd=apply(x.c, 1, sd)) # return
answer
+ })
> # show first element
> head(x.ans[[1]])
MODEL_1 MODEL_10 MODEL_11 MODEL_12 MODEL_13 MODEL_14 MODEL_15
MODEL_16 MODEL_17 MODEL_18 MODEL_19 MODEL_2 MODEL_20
1_ALA 84.47 83.65 35.68 90.86 5.22 25.07 83.06
91.91 94.22 26.01 86.16 33.59 96.73
10_THR 112.37 111.48 107.58 115.86 111.25 115.27 112.66
115.42 107.43 108.04 116.90 116.78 122.04
100_GLN 78.52 66.41 71.85 86.28 67.72 84.80 74.74
89.17 86.58 72.23 100.27 89.84 84.64
101_THR 88.12 75.22 76.51 79.97 70.00 76.24 77.20
62.92 79.20 77.10 69.78 85.96 78.51
102_ILE 127.82 125.73 128.98 125.70 125.30 127.42 126.74
130.55 124.48 130.89 123.61 126.51 129.49
103_ARG 23.91 34.13 57.48 16.20 87.60 22.04 34.82
79.69 37.68 42.45 25.67 44.80 35.99
MODEL_3 MODEL_4 MODEL_5 MODEL_6 MODEL_7 MODEL_8 MODEL_9
mean sd
1_ALA 64.76 17.03 64.30 99.59 10.69 19.42 40.83 57.6625
33.499007
10_THR 110.82 115.20 111.98 119.53 104.27 112.14 110.21 112.8615
4.346347
100_GLN 78.79 90.64 71.63 81.45 65.87 69.15 63.74 78.7160
10.122561
101_THR 77.12 68.42 74.01 59.58 61.30 76.27 87.44 75.0435
7.909933
102_ILE 126.79 123.19 129.90 122.54 125.35 131.20 130.78 127.1485
2.715634
103_ARG 11.06 9.83 17.08 41.61 28.51 44.08 14.14 35.4385
20.886373
On 2/28/07, Aimin Yan <[EMAIL PROTECTED]> wrote:
> In order to make this question be clear, Here is my code
>
> >source("http://omega.psi.iastate.edu/data/nmr_19.R";)
> > class(temp)
> [1] "list"
> > names(temp)
> [1] "1A24" "1A57" "1A5J" "1A6X" "1AB7" "1AF8" "1AFI" "1AGG" "1AH9" "1AHL"
> [11] "1AJ3" "1AJW" "1AK7" "1ALG" "1AOY" "1APF" "1AUZ" "1AZJ" "1AZK"
> > head(temp[[1]])
> pr model aa omega
> 1 1A24 MODEL_1 1_ALA 84.47
> 2 1A24 MODEL_1 2_GLN 63.06
> 3 1A24 MODEL_1 3_TYR 107.72
> 4 1A24 MODEL_1 4_GLU 54.36
> 5 1A24 MODEL_1 5_ASP 67.01
> 6 1A24 MODEL_1 6_GLY 999.00
>
> This Iist have 19 data.frame that has their name such as 1A24,1A57....
> The above is the structure for data.frame 1A24, pr have 1 level,
> model has 20 levels,aa has 189 levels
>
> I want to get the following matrix for each data.frame in this list like
> this
>
> 1A24
> MODEL_1 MODEL_2 ..... MODEL_20 Mean_of_20_models
> Sd_of_20_models
> 1_ALA 84.47
> 2_GLN 63.06
> 3_TYR 107.72
> 4_GLU 54.36
> 5_ASP
> 6_GLY
>
> same thing for other data.frame in this list
>
> 1A57
> .....
>
> 1A5J
> .....
>
> Means_of_20_models is average of omega for 1_ALA in 20 models
> Sd_of_20_models is standard deviation of omega for 1_ALA between 20 models
>
>
> Does anyone has some advice on how to do this?
>
> thanks
>
> Aimin
>
> ______________________________________________
> [email protected] 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.
>
--
Jim Holtman
Cincinnati, OH
+1 513 646 9390
What is the problem you are trying to solve?
[[alternative HTML version deleted]]
______________________________________________
[email protected] 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.
