Error using H2O xgboost model about lime HOT 4 CLOSED

This doesn't really have anything to do with lime - if you run:

h2o.predict(full_mdl, as.h2o(dplyr::select(as.data.frame(full_iris_frame)[1:4,], -Species)))

you get the same error. I'm afraid I can help you more here as my experience with H2O is rather low. Maybe @mdancho84 can chime in..?

from lime.

@dkincaid - Let me work @DVaughan on this. For some reason h2o.xgboost is not available on Windows, and I’m a PC guy (although probably switching).

@thomasp85 - your going to be getting a bunch of H2O questions. I’ll be speaking at H2O World in December about your pkg. Just an FYI.

from lime.

I think it would be beneficial to move this discussion to the H2O repo as it is no longer lime related (I simply thought @mdancho84 had a quick resolution)

from lime.

I agree. Not a simple solution, and it likely pertains to the h2o.xgboost algorithm.

from lime.