Partial canonical analysis is the extension of partial linear regression to the case where the response data table Y is multivariate. The logic is the same as explained in Subsection 10.3.5 and will not be detailed further. Partial canonical analysis includes partial RDA (Davies & Tso, 1982) and partial CCA (ter Braak, 1988a). As in partial linear regression, the explanatory variables are seen as belonging to two groups: X for the explanatory variables to be included in the model and W for the covariables (also
Was this article helpful?