API¶

Main Function¶

The main function provided by the package formulates a predictor in a Gurobi model. In most cases, it is the only function needed from the package.

Some formulations of the predictors can have additional options. These are documented in the specific functions for each regression model.

The main entry point gurobi_ml.add_predictor_constr() also accepts advanced keyword arguments (verbose, no_debug, no_record) that are forwarded to internal formulation builders.

add_predictor_constr(gp_model, predictor, ...)

Formulate predictor in gp_model.

Scikit-learn API¶

`column_transformer`	Module for formulating a `sklearn.compose.ColumnTransformer` in a `gurobipy model`.
`decision_tree_regressor`	Module for formulating a `sklearn.tree.DecisionTreeRegressor` in a `Model`.
`gradient_boosting_regressor`	Module for formulating a `sklearn.ensemble.GradientBoostingRegressor` into a `Model`.
`linear_regression`	Module for formulating ordinary regression models into a `Model`.
`logistic_regression`	Module for formulating a `sklearn.linear_model.LogisticRegression` in a `Model`.
`mlpregressor`	Module for formulating a `sklearn.neural_network.MLPRegressor` in a `Model`.
`pipeline`	Module for formulating a `sklearn.pipeline.Pipeline` in a `Model`.
`pls_regression`	Module for formulating `sklearn.cross_decomposition.PLSRegression` in a gurobipy model.
`random_forest_regressor`	Module for formulating a `sklearn.ensemble.RandomForestRegressor` into a `Model`.
`preprocessing`	Implementation for using sklearn preprocessing object in a Gurobi model.