Model development: Annotating Categorical Variables

[kzollove]

The problem as I understand it is that categorical variables are transferred from the L0 to the L1 dataset, but the definitions of those variable remain in the L0 metadata.

Annotating categorical variables of the L0, which become values in L1, would facilitate full transfer of the L0 content to the L1, which negates the effort to lookup the content in the L0.

This could be implemented by adding a “value” column to variable_mapping table.