This issue is related to #58. It would be nice to also obtain the confidence intervals for the weights instead of the loadings. For my project I am more interested in analyzing the weights instead of the loadings because they take multicolinearity into account.