[R] cor() btwn columns in two matrices - no complete element pairs

rcoder mpdotbook at gmail.com
Fri Aug 15 16:13:24 CEST 2008

Is there any way to increment the loop when the error occurs, returning an NA
in the results matrix. I have included an if starement to only calculate
cor() when columns are !all(is.na()), but I still exit out with this error,
albeit after some loops through the matrix - i.e. when the condition is
satisfied, but the data rows do not overlap at all.

rcoder wrote:
> Hi everyone,
> I'm trying to calculate correlation coefficients between corresponding
> columns in two matrices with identical dimensions but different data. The
> problem is that the matrices contain NAs in different locations. I am
> using the following code to try to calculate correlations between complete
> sets of data:
> #Code start
> maxcol<-ncol(mat1)
> for (i in 1:maxcol)
> 	{
> corr_results[1,i]<-cor(mat1[,i],mat2[,i], use="complete.obs")
> 	}
> #Code end
> ...but I get the following error message:
> Error in cor(mat1[,i], mat2[,i], use="complete.obs") : 
>   no complete element pairs
> Is there something I'm not including in the 'cor' parentheses? I apologise
> for not including the true original data frames.
> Thanks,
> rcoder

View this message in context: http://www.nabble.com/cor%28%29-btwn-columns-in-two-matrices---no-complete-element-pairs-tp18998875p18999356.html
Sent from the R help mailing list archive at Nabble.com.

More information about the R-help mailing list