Thread Subject:

Correlation co-efficient between two matrices of different order

Hi,
I need to:
i) calculate a matrix of correlation coefficients between the column vectors of two matrices, of which A=600*750 dimensions and B=600*830 (the rows are same variables but columns are different variables), and
ii) from the resulting matrix of correlation co-efficients, extract another matrix wherein those elements are set to be 1 which were significantly positively correlated at p<0.01, and 0 otherwise.

Any help will be highly appreciated.

Best regards,
Shafi

"Muhammad Shafique" wrote in message <jc4s0p$3c$1@newscl01ah.mathworks.com>...
> Hi,
> I need to:
> i) calculate a matrix of correlation coefficients between the column vectors of two matrices, of which A=600*750 dimensions and B=600*830 (the rows are same variables but columns are different variables), and
> ii) from the resulting matrix of correlation co-efficients, extract another matrix wherein those elements are set to be 1 which were significantly positively correlated at p<0.01, and 0 otherwise.
>
> Any help will be highly appreciated.
>
> Best regards,
> Shafi
- - - - - - - - - -
  You could use corrcoef([A,B]) and then simply discard the unwanted correlations. Perhaps use the second returned argument to obtain the p-values.

Roger Stafford