I have a compositional data and need to use compositional data analysis? The independent variables are ratios and hence I cannot use regular classical regression without variable transformations.
y=A + BM + CN + error; where M, N and O are ratios such that M+N+O=1. A,B and C are model parameters. Y is a continuous variable. The parameters have to estimated with the constraint that the sum of the ratios=1.
Can this be done in SAS?
How do you transform a variable using isometric log-ratio (ILR) transformation using SAS?
Any help would be greatly appreciated.
question from:https://stackoverflow.com/questions/65647271/compositional-data-and-isometric-log-ratio-ilr-transformation-in-sas
