Multicategorical antecedent X variables using mcx=1?

[WestfallDaniel]

Does this package support multicategorical antecedent X variables such as in the Hayes Process program using mcx=1?

Thank you!

[QuentinAndre]

Dear Daniel,

Unfortunately, not as of now. It would be a good idea for a pull request, but do not have time to work on it in the near future. I will keep you posted if I add this feature!

Best,

Quentin.

[QuentinAndre]

Scheduling this for the 2.1.0 milestone as the multicategorical-predictor feature.

Scope. PROCESS 3's mcx and mcw options: indicator (dummy) coding, sequential coding and Helmert coding for a multicategorical X and, where the model allows, a multicategorical W. Relative direct and indirect effects per indicator, omnibus tests where PROCESS reports them, and the categorical variable accepted as pandas.Categorical, strings, or integers (which also covers #24's request to cast automatically).

Tests. PROCESS 3 fixtures for at least model 4 and model 7 with a three-level X; accuracy suite extended.

[QuentinAndre]

Moving this from 2.1.0 to 2.2.0. The rest of 2.1 is implemented; this feature needs PROCESS 3/4 comparison output to be validated (indicator, sequential and Helmert coding; relative direct and indirect effects; omnibus tests), which #82 sets up, and it changes how X is represented throughout the symbolic-derivative code. Shipping it unverified would be worse than shipping it later.