Closed Deleetdk closed 2 years ago
I am getting the same error for one of my analyses. Do you manage to resolve it? Or omitting the problematic variable is the only option?
I am getting the same error for one of my analyses. Do you manage to resolve it? Or omitting the problematic variable is the only option?
I was only doing this dataset for testing purposes, I omitted it. https://rpubs.com/EmilOWK/BMA_examples
This error in most cases points to collinearity. In the case of the mpg dataset however, something else is going on, maybe related to the fact that the "model" variable has 38 levels. Even when running glm on the original dataset, 14 out of the 37 "model" related coefficients cannot be identified. So as in the case of collinearity, omitting the variable is the way to go.
I found an error. Not sure of the cause, but relates to one particular variable (model). Here's a reprex. I guess it may be due to handling factor variable with too many small levels.
Created on 2021-01-18 by the reprex package (v0.3.0.9001)