[R] Howto Compute Pairwise Similarity/Correlation Matrix from a Data Frame

Gundala Viswanath gundalav at gmail.com
Wed Jun 18 08:55:48 CEST 2008


Hi,

I have the following 5 vectors. I wish to compute
the pairwise Pearson Correlation matrix with this data.frame.
Is there a compact way to do it?

At the end I hope to create a heatmap out of this correlation matrix.

__BEGIN__
> data <- read.table("mydata.txt")
> print(data)
          V1      V2      V3      V4      V5      V6      V7      V8      V9
1       42.3    53.2    76.4    78.8    83.6    91.3    92.2   105.8   109.6
2        6.8     9.7    12.7    13.1    14.6    16.3    17.2    17.9    18.1
3       10.6    21.5    34.4    38.2    38.8    50.0    60.7    64.0    64.3
4      215.3   227.4   227.7   245.0   257.2   260.0   269.8   287.3   340.2
5        4.1     4.2     4.6     6.4     6.8     6.9    16.9    17.6    23.3

__END__

Currently I am stuck in constructing the very matrix itself from
double loop.

__BEGIN__
data <- read.table("GDS596_part1.txt")
nofrow <- nrow(data)

for (rwx in 1:nofrow) {
   print(data[rwx,])
   for (rwy in 1:nofrow) {
      print(data[rwy,])

       thecor <- cor(data[rwx,],data[rwy.], method="pearson")
       # not sure how to proceed from here.
   }

}
__END__

Please advice.

- Gundala Viswanath
Jakarta - Indonesia



More information about the R-help mailing list