Hi Wati,

Canonical analysis can be a powerful tool if used properly. The Eigen values tell us the amount of variance that each dimension of the dataset accounts for. By summing the Eigen values and dividing each Eigen value by that sum, you can estimate the proportion of variance that dimension accounts for. More than two or three dimensions are seldom meaningful in a canonical analysis. To help you with you questions about presenting the results, I am going to attach two papers. These demonstrate its use, presentation, and interpretation.

Here is the second paper:

James E. McLean thank you for the references sir, i shall look into them. Regards