transformation of data.frame

Hello Petr,

sorry for the mixed up. your example works perfectly fine. 

The one from S?ren has shown the mentioned error.  But even after 

columns as character

go <- read.table("go.txt", header= TRUE, colClasses = c("character", 

or

go <- read.table("go.txt", header= TRUE, as.is = 1)

it didn't solve the problem.
the command:
gmt <- lapplyBy(~GO, data = go, FUN = function(uu) {as.list(uu$GO[1], 

$gen, collapse = " "))})

tries to convert my first column into integers and thand add 'NA's. 

What I don't understand is why. 
Does lapplyBy can work only with integers?

THX,

Assa

2010/7/8 Petr PIKAL <petr.pikal at precheza.cz>
Hi

r-help-bounces at r-project.org napsal dne 08.07.2010 12:02:45:

I don't understand it. When I'm doing this example it wirks fine, but

when

I'm adding the "GO:" to the beginning of the first column (as to see 

my

wanted result table:
GO0042787
GO0016070
GO0016070

I'm getting a list of warning:
Warning messages:
1: In storage.mode(xi) <- a$sm : NAs introduced by coercion
2: In storage.mode(xi) <- a$sm : NAs introduced by coercion
...
9: In storage.mode(xi) <- a$sm : NAs introduced by coercion
10: In storage.mode(xi) <- a$sm : NAs introduced by coercion

Not sure what is wrong, it seems to me that your ID become factor.

Having your data in dataframe test as character columns

see ?str

test.ag<-aggregate(test$X.gen, list(test$ID), function(x) paste(x,
collapse=":"))

I can make aggregated data frame

paste("GO",test.ag[,1], sep="")
[1] "GO0006417" "GO0006511" "GO0007409" "GO0016070" "GO0042787"

and it is strightforward to add GO at the beginning.

I leave how to add this result to your aggregated data frame as an
exercise.

Regards
Petr

What did I do wrong here?

Assa

On Thu, Jul 8, 2010 at 11:09, S??ren H??jsgaard

<Soren.Hojsgaard at agrsci.dk>wrote:

Like this?

library(doBy)
(ddd <- read.table("foo.txt",header=T))

    ID  gen
1 42787 gen2
2 16070 gen2
3 16070 gen3
4  7409 Gen1
5  7409 gen3
6  6511 gen2
7  6417 gen3
8 16070 gen4
9  6511 gen4

aa<-lapplyBy(~ID, data=ddd,

+   FUN=function(uu){
+   list(uu$ID[1], paste(uu$gen, collapse=":"))
+ })

do.call(rbind,aa)

     [,1]  [,2]
42787 42787 "gen2"
16070 16070 "gen2:gen3:gen4"
7409  7409  "Gen1:gen3"
6511  6511  "gen2:gen4"
6417  6417  "gen3"

Regards
S??ren





-----Oprindelig meddelelse-----
Fra: r-help-bounces at r-project.org [

] P??

vegne af Assa Yeroslaviz
Sendt: 8. juli 2010 10:45
Til: r-help at stat.math.ethz.ch
Emne: [R] transformation of data.frame

Hello all R users,

I have a problems transforming (or maybe better regrouping) a

data.frame.

 I have a big data.frame, which I would like to sum up according to 

specific column.

This is an example of my matrix:
ID    gen
0042787    gen2
0016070    gen2
0016070    gen3
0007409    Gen1
0007409    gen3
0006511    gen2
0006417    gen3
0016070    gen4
0006511    gen4

I want to rearrange the matrix according to column GO, so that it 

look

likes that:

GO:0042787     gen2
GO:0016070    gen2  :  gen3  :  gen4
GO:0007409    gen1  :  gen3
GO:0006511    gen2  :  gen4
GO:0006417    gen3

I've tried it with the package doBy (lapplyBy and paste) but it just
doesn't
work out.

I will be very happy for any suggestions you might have to help me.

Thanks

Assa

        [[alternative HTML version deleted]]

   [[alternative HTML version deleted]]