10.10.0

Author: mathias-von-ottenbreit

API_REFERENCE_FOR_REGRESSION.md

@@ -368,4 +368,8 @@ Prevents the output from having significantly more than ***max_rows_before_sampl
 ### Parameters
 
 #### value
-A float representing the new intercept.
+A float representing the new intercept.
+
+## Method: remove_provided_custom_functions()
+
+***Removes any custom functions provided for calculating the loss, negative gradient, or validation error. This is useful after model training with custom functions, ensuring that the APLRRegressor object no longer depends on these functions—so they do not need to be present in the Python environment when loading a saved model.***

aplr/aplr.py

@@ -312,6 +312,12 @@ def get_cv_error(self) -> float:
     def set_intercept(self, value: float):
         self.APLRRegressor.set_intercept(value)
 
+    def remove_provided_custom_functions(self):
+        self.APLRRegressor.remove_provided_custom_functions()
+        self.calculate_custom_validation_error_function = None
+        self.calculate_custom_loss_function = None
+        self.calculate_custom_negative_gradient_function = None
+
     # For sklearn
     def get_params(self, deep=True):
         return {

cpp/APLRRegressor.h

@@ -304,6 +304,7 @@ class APLRRegressor
                                                         size_t unique_term_affiliation_index);
     double get_cv_error();
     void set_intercept(double value);
+    void remove_provided_custom_functions();
 
     friend class APLRClassifier;
 };
@@ -2775,4 +2776,11 @@ void APLRRegressor::set_intercept(double value)
         throw std::runtime_error("The new intercept must be finite.");
     intercept = value;
     term_coefficients[0] = value;
+}
+
+void APLRRegressor::remove_provided_custom_functions()
+{
+    calculate_custom_validation_error_function = {};
+    calculate_custom_loss_function = {};
+    calculate_custom_negative_gradient_function = {};
 }

cpp/pythonbinding.cpp

@@ -80,6 +80,7 @@ PYBIND11_MODULE(aplr_cpp, m)
              py::arg("max_rows_before_sampling") = 100000)
         .def("get_cv_error", &APLRRegressor::get_cv_error)
         .def("set_intercept", &APLRRegressor::set_intercept, py::arg("value"))
+        .def("remove_provided_custom_functions", &APLRRegressor::remove_provided_custom_functions)
         .def_readwrite("intercept", &APLRRegressor::intercept)
         .def_readwrite("m", &APLRRegressor::m)
         .def_readwrite("m_optimal", &APLRRegressor::m_optimal)

cpp/tests.cpp

@@ -661,7 +661,7 @@ class Tests
         model.fit(X_train, y_train, sample_weight, {}, cv_observations, prioritized_predictor_indexes, {}, VectorXi(0), {}, X_train);
         std::cout << "feature importance\n"
                   << model.feature_importance << "\n\n";
-
+        model.remove_provided_custom_functions();
         VectorXd predictions{model.predict(X_test)};
 
         // Saving results

documentation/APLR 10.9.0.pdf renamed to documentation/APLR 10.10.0.pdf

@@ -28,7 +28,7 @@
 
 setuptools.setup(
     name="aplr",
-    version="10.9.0",
+    version="10.10.0",
     description="Automatic Piecewise Linear Regression",
     ext_modules=[sfc_module],
     author="Mathias von Ottenbreit",