in SLIM, differentiate between the case when user is missing cvxpy and the case when a user has cvxpy but doesn't have a powerful enough solver for mixed-integer linear or logistic regression
make DecisionTreeClassifierWithComplexity the backup model when user doesn't have gosdt installed and provide a warning message
no longer conditionally exclude imodels.tree.gosdt from the built package based on OS
cvxpy
and the case when a user hascvxpy
but doesn't have a powerful enough solver for mixed-integer linear or logistic regressionimodels.tree.gosdt
from the built package based on OS