Would be great if you took at the categorical approach to make sure it looks like the other approaches.
One question I had was with passing an argument to explain. Should this be accessed later through internal$parameters or some other way? (Couldn't find the answer looking through the Gaussian method).
No tests yet but I manually checked that the results give the same output as the old shapr package.
No need to merge the branch until the tests are done.
Hi Martin,
Would be great if you took at the categorical approach to make sure it looks like the other approaches.
One question I had was with passing an argument to explain. Should this be accessed later through internal$parameters or some other way? (Couldn't find the answer looking through the Gaussian method).
No tests yet but I manually checked that the results give the same output as the old shapr package.
No need to merge the branch until the tests are done.
Annabelle