Hi there,

I would like to report a bug that occurred when upgrading from 0.23.1 to 0.24.1. When fitting the CCA models in 0.23.2, everything appeared to be working as planned and was able to find many correspondent predictors when investigating the coefficients. The projections also appeared to demonstrate nice variability amongst samples when projecting to 2-d. However, when upgrading to 0.24.1, I noticed that CCA appeared to drop any additional axes outside of the first axis, in other words, the loadings for any axis outside of the first would converge to zero. I'm not sure why, as variation via common predictors was well accounted for using 0.23.2. I know there were upgrades that were done on CCA between 0.23 and 0.24 that improved stability. I downgraded back to 0.23 and all seems well. Just thought I'd post here to see if anyone else is experiencing this issue.

Thanks!