[R] R Error : DATA to MATRIX

ritwik_r at isical.ac.in ritwik_r at isical.ac.in
Fri Mar 23 07:53:56 CET 2012


Dear Sir/Madam,

I'm getting a problem with a R-code which converts a data frame to a matrix.

It first generate a (m^(n-m) * m) matrix A and then regenerate another
matrix B having less dimension than A which satisfy some condition. Now I
wish to assign each row of B to a vector as individual.

My problem is when I set any choice of (n,m) except m=1 it works fine but
setting m=1 I got the error : Error in B[i, ] : incorrect number of
dimensions.

Moreover if (n,m) is large (say, (20,8)) I got the error : Error: cannot
allocate vector of size 3.0 Gb. I know this is due to large dimension of
matrix A. How to solve this problem.

My code is given below:

**********************************************************************

n=5
m=3
R=numeric(0)
# Generate all possible m-tuple ( variables having range 0 to n  ) in a (
m^(n-m) * m ) matrix

r = expand.grid(rep(list(0:(n-m)), m))

write.table(r,file="test.txt",row.names=FALSE,col.names=FALSE)

a= read.table(file="test.txt",sep="",header=FALSE)

A= data.matrix(a)

#.........................................................................................

# Generate matrix whose rowsum = n-m

meet.crit = apply(A, 1, function(.row) any((sum(.row)) == n-m))      # 
criteron for being rowsum = n

cbind(A, meet.crit)                                                  # 
Checking rowsum = n for each row
-m
B=A[meet.crit,]                                                      # 
Generate matrix

#.........................................................................................


for(i in 1:choose(n-1,m-1)){
R=B[i,]
}

***************************************************************************

Can you please help me how to get rid of these errors. Thanking you in
advance.

Regards

Ritwik Bhattacharya


Senior Research Fellow
SQC & OR UNIT, KOLKATA
INDIAN STATISTICAL INSTITUTE

Voice : +91 9051253944

This mail is scanned by Ironport



More information about the R-help mailing list