R help - Dec 2004 - numeric(0)

Dear all,

I am trying to calculate a score for a string sequence consisting of 
the following four letters: ACGT.
I have got a matrix giving the scores for each pair of letters.
So for example the string ACCT has got the pairs: AC, CC and CT.

The matrix has got the following form:
names<-c("A","C","G","T");
mscore<-matrix(0,4,4);
rownames(mscore)<-names;
colnames(mscore)<-names;

So for the first example pair above I could get the score contained in 
the matrix with the following code:
 >> mscore["A","C"]

I am now trying to sum up all the scores with the following code:

score<-0;

for(j in 1:length(sequence)-1){
	score<-score+mscore[sequence[j],sequence[j+1]];
}

where sequence is the string sequence.
Unfortunately, it does not work and I get the following error message:

Error in "[<-"(`*tmp*`, 1, i, value = numeric(0)) :
	nothing to replace with

What does this mean? Strangely, the command
"print(score+mscore[sequence[j],sequence[j+1]])"
works, so it really is the assignment which won't work. Why is that?
I am running R 2.0 series (GUI version) on Mac OSX 10.3.7.

Any suggestions are welcome.
Thanks a lot,
Dax

On 29-Dec-04 dax42 wrote:
> I am now trying to sum up all the scores with the following code:
> 
> score<-0;
> 
> for(j in 1:length(sequence)-1){
>       score<-score+mscore[sequence[j],sequence[j+1]];
> }
> 
> where sequence is the string sequence.
> Unfortunately, it does not work and I get the following error message:
> 
> Error in "[<-"(`*tmp*`, 1, i, value = numeric(0)) :
>       nothing to replace with
You've fallen into a classic trap: 1:length(sequence)-1 does not
mean what one might naturally expect.
> sequence<-(1:10)
> length(sequence)
[1] 10
> 1:length(sequence)-1
 [1] 0 1 2 3 4 5 6 7 8 9
> 1:(length(sequence)-1)
[1] 1 2 3 4 5 6 7 8 9

In other words, "1:length(sequence)" is constructed first, then
1 is subtracted from every element. As a result, you tried to
read from sequence[0], which isn't there.

However, you can make it evaluate "length(sequence)-1" before
constructing 1:(length(sequence)-1), by using the parantheses
to force precedence.

Cheers,
Ted.


--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 094 0861  [NB: New number!]
Date: 29-Dec-04                                       Time: 13:45:25
------------------------------ XFMail ------------------------------

You would probably be better off without the loop, for example:

ns <- length( sequence )
num.seq <- match( sequence, names )
scores <- mscore[cbind(num.seq[-1],num.seq[-ns])]
sum( scores )

I have used the fact that if you index a matrix with
a two-column,matrix ( here, cbind( , ) ), you select the
corresponding elements of the matrix, see ?"["
But it only works if the matrix is numeric.

Bendix Carstensen
----------------------
Bendix Carstensen
Senior Statistician
Steno Diabetes Center
Niels Steensens Vej 2
DK-2820 Gentofte
Denmark
tel: +45 44 43 87 38
mob: +45 30 75 87 38
fax: +45 44 43 07 06
bxc at steno.dk
www.biostat.ku.dk/~bxc
----------------------
> -----Original Message-----
> From: r-help-bounces at stat.math.ethz.ch 
> [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of dax42
> Sent: Wednesday, December 29, 2004 2:18 PM
> To: r-help at stat.math.ethz.ch
> Subject: [R] numeric(0)
> 
> 
> Dear all,
> 
> I am trying to calculate a score for a string sequence consisting of 
> the following four letters: ACGT.
> I have got a matrix giving the scores for each pair of 
> letters. So for example the string ACCT has got the pairs: 
> AC, CC and CT.
> 
> The matrix has got the following form: 
> names<-c("A","C","G","T");
mscore<-matrix(0,4,4);
> rownames(mscore)<-names; colnames(mscore)<-names;
> 
> So for the first example pair above I could get the score 
> contained in 
> the matrix with the following code:
>  >> mscore["A","C"]
> 
> I am now trying to sum up all the scores with the following code:
> 
> score<-0;
> 
> for(j in 1:length(sequence)-1){
> 	score<-score+mscore[sequence[j],sequence[j+1]];
> }
> 
> where sequence is the string sequence.
> Unfortunately, it does not work and I get the following error message:
> 
> Error in "[<-"(`*tmp*`, 1, i, value = numeric(0)) :
> 	nothing to replace with
> 
> What does this mean? Strangely, the command 
> "print(score+mscore[sequence[j],sequence[j+1]])"
> works, so it really is the assignment which won't work. Why 
> is that? I am running R 2.0 series (GUI version) on Mac OSX 10.3.7.
> 
> Any suggestions are welcome.
> Thanks a lot,
> Dax
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list 
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read 
> the posting guide! http://www.R-project.org/posting-guide.html
>

On Dec 29, 2004, at 8:17 AM, dax42 wrote:
> Dear all,
>
> I am trying to calculate a score for a string sequence consisting of 
> the following four letters: ACGT.
> I have got a matrix giving the scores for each pair of letters.
> So for example the string ACCT has got the pairs: AC, CC and CT.
>
> The matrix has got the following form:
> names<-c("A","C","G","T");
> mscore<-matrix(0,4,4);
> rownames(mscore)<-names;
> colnames(mscore)<-names;
>
> So for the first example pair above I could get the score contained in 
> the matrix with the following code:
> >> mscore["A","C"]
>
> I am now trying to sum up all the scores with the following code:
>
> score<-0;
>
> for(j in 1:length(sequence)-1){
> 	score<-score+mscore[sequence[j],sequence[j+1]];
> }
>
Is sequence a string?  If so, you will probably need to make some 
modifications like these:

 > sequence <- 'ACGT'
 > sequence[1]
[1] "ACGT"
 > substr(sequence,1,1)
[1] "A"
 > substr(sequence,2,1)
[1] ""
 > substr(sequence,1,2)
[1] "AC"
 > substr(sequence,2,2)
[1] "C"
 > substr(sequence,3,3)
[1] "G"
 > mscore <- matrix(runif(16),4,4)
 > names <- c('A','C','G','T')
 > rownames(mscore) <- names
 > colnames(mscore) <- names
 > mscore
           A         C         G          T
A 0.6200289 0.6324337 0.1895207 0.28253473
C 0.5026072 0.6552428 0.7978809 0.43131540
G 0.1669823 0.8808445 0.6021024 0.01563101
T 0.4184646 0.9620714 0.7723088 0.33045464
 > mscore[substr(sequence,1,1),substr(sequence,2,2)]
[1] 0.6324337
 > score <- 0
 > for (j in 1:(nchar(sequence)-1)) {score <- score+ 
mscore[substr(sequence,j,j),substr(sequence,j+1,j+1)]}
 > score
[1] 1.445946

Hope this helps.  I imagine this isn't the most efficient way of 
solving this problem, though.

Sean

`names' is a builtin R function for extracting names of elements in a
vector, or names of components in a list.  If I use

nm <- c("A","C","G","T")
rownames(mscore) <- colnames(mscore) <- nm

I don't get any error.

Andy
> From: dax42
> 
> Dear all,
> 
> I am trying to calculate a score for a string sequence consisting of 
> the following four letters: ACGT.
> I have got a matrix giving the scores for each pair of letters.
> So for example the string ACCT has got the pairs: AC, CC and CT.
> 
> The matrix has got the following form:
> names<-c("A","C","G","T");
> mscore<-matrix(0,4,4);
> rownames(mscore)<-names;
> colnames(mscore)<-names;
> 
> So for the first example pair above I could get the score 
> contained in 
> the matrix with the following code:
>  >> mscore["A","C"]
> 
> I am now trying to sum up all the scores with the following code:
> 
> score<-0;
> 
> for(j in 1:length(sequence)-1){
> 	score<-score+mscore[sequence[j],sequence[j+1]];
> }
> 
> where sequence is the string sequence.
> Unfortunately, it does not work and I get the following error message:
> 
> Error in "[<-"(`*tmp*`, 1, i, value = numeric(0)) :
> 	nothing to replace with
> 
> What does this mean? Strangely, the command
> "print(score+mscore[sequence[j],sequence[j+1]])"
> works, so it really is the assignment which won't work. Why is that?
> I am running R 2.0 series (GUI version) on Mac OSX 10.3.7.
> 
> Any suggestions are welcome.
> Thanks a lot,
> Dax
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! 
> http://www.R-project.org/posting-guide.html
> 
>

> 
: Date:   Wed, 29 Dec 2004 14:17:35 +0100 
: From:   dax42 <Dax42 at web.de>
: [ Add to Address Book | Block Address | Report as Spam ] 
: To:   <r-help at stat.math.ethz.ch> 
: Subject:   [R] numeric(0) 
: 
:  
: Dear all,
: 
: I am trying to calculate a score for a string sequence consisting of 
: the following four letters: ACGT.
: I have got a matrix giving the scores for each pair of letters.
: So for example the string ACCT has got the pairs: AC, CC and CT.
: 
: The matrix has got the following form:
: names<-c("A","C","G","T");
: mscore<-matrix(0,4,4);
: rownames(mscore)<-names;
: colnames(mscore)<-names;
: 
: So for the first example pair above I could get the score contained in 
: the matrix with the following code:
: >> mscore["A","C"]
: 
: I am now trying to sum up all the scores with the following code:
: 
: score<-0;
: 
: for(j in 1:length(sequence)-1){
:      score<-score+mscore[sequence[j],sequence[j+1]];
: }

The above code sutracts 1 from the vector 1:length(sequence) giving
a vector that starts at 0.  I think you meant 

	1:(length(sequence)-1)
: 
: where sequence is the string sequence.
: Unfortunately, it does not work and I get the following error message:
: 
: Error in "[<-"(`*tmp*`, 1, i, value = numeric(0)) :
:      nothing to replace with
: 
: What does this mean? Strangely, the command
: "print(score+mscore[sequence[j],sequence[j+1]])"
: works, so it really is the assignment which won't work. Why is that?
: I am running R 2.0 series (GUI version) on Mac OSX 10.3.7.
: 
: Any suggestions are welcome.

Try this.  (The first line is to ensure that levels that don't
appear still get counted.)

f <- factor(sequence, levels = names)
sum(table(f[-length(f)], f[-1]) * mscore)

From:   <Ted.Harding at nessie.mcc.ac.uk>
> What's the best place to look for the details on operator
> precedence and the like in R?
Check out the language reference manual in the directory
   .../R/rw2001/doc/manual/r-lang.pdf .

On Windows its also accessible from R via 
   Help | Manuals | R Language

The manuals can also be found online.  e.g. Google for 
   r-lang "infix and prefix"
to find the relevant section.

By the way, even without this problem seq is better than : in 
programs as it has better boundary behavior.  Consider 1:n vs.
seq(length = n).  For n=0 the : operator actually generates the
vector c(1,0) whereas the seq solution gives a zero length
vector, as desired.  Even seq has problems unless you use
length= .

The : operator is mainly useful if you are typing throwaway 
code directly into R.