aplr 10.17.1__cp310-cp310-win32.whl → 10.18.1__cp310-cp310-win32.whl

aplr/aplr.py

@@ -1,5 +1,6 @@
 from typing import List, Callable, Optional, Dict, Union
 import numpy as np
+import pandas as pd
 import aplr_cpp
 import itertools
 
@@ -9,7 +10,167 @@ IntVector = np.ndarray
 IntMatrix = np.ndarray
 
 
-class APLRRegressor:
+class BaseAPLR:
+    def _validate_X_fit_rows(self, X):
+        """Checks if X has enough rows to be fitted."""
+        if (isinstance(X, np.ndarray) and X.shape[0] < 2) or (
+            isinstance(X, pd.DataFrame) and len(X) < 2
+        ):
+            raise ValueError("Input X must have at least 2 rows to be fitted.")
+
+    def _common_X_preprocessing(self, X, is_fitting: bool, X_names=None):
+        """Common preprocessing for fit and predict."""
+        is_dataframe_input = isinstance(X, pd.DataFrame)
+
+        if X_names is not None:
+            X_names = list(X_names)
+
+        if not is_dataframe_input:
+            try:
+                X_numeric = np.array(X, dtype=np.float64)
+            except (ValueError, TypeError) as e:
+                raise TypeError(
+                    "Input X must be numeric if not a pandas DataFrame."
+                ) from e
+            X = pd.DataFrame(X_numeric)
+            if is_fitting:
+                if X_names:
+                    X.columns = X_names
+                else:
+                    X.columns = [f"X{i}" for i in range(X.shape[1])]
+            elif self.X_names_ and len(self.X_names_) == X.shape[1]:
+                X.columns = self.X_names_
+        else:  # X is already a DataFrame
+            X = X.copy()  # Always copy to avoid modifying original
+            if not is_fitting and self.X_names_:
+                # Check if input columns for prediction match training columns (before OHE)
+                if set(X.columns) != set(self.X_names_):
+                    raise ValueError(
+                        "Input columns for prediction do not match training columns."
+                    )
+                X = X[self.X_names_]  # Ensure order of original columns
+
+        if is_fitting:
+            self.X_names_ = list(X.columns)
+            self.categorical_features_ = list(
+                X.select_dtypes(include=["category", "object"]).columns
+            )
+            # Ensure it's an empty list if no categorical features, not None
+            if not self.categorical_features_:
+                self.categorical_features_ = []
+
+        # Apply OHE if categorical_features_ were found during fitting.
+        if self.categorical_features_:
+            X = pd.get_dummies(X, columns=self.categorical_features_, dummy_na=False)
+            if is_fitting:
+                self.ohe_columns_ = list(X.columns)
+                # Ensure it's an empty list if no OHE columns, not None
+                if not self.ohe_columns_:
+                    self.ohe_columns_ = []
+            else:
+                missing_cols = set(self.ohe_columns_) - set(X.columns)
+                for c in missing_cols:
+                    X[c] = 0
+                X = X[self.ohe_columns_]  # Enforce column order
+
+        if is_fitting:
+            self.na_imputed_cols_ = [col for col in X.columns if X[col].isnull().any()]
+            # Ensure it's an empty list if no NA imputed columns, not None
+            if not self.na_imputed_cols_:
+                self.na_imputed_cols_ = []
+
+        # Apply NA indicator if na_imputed_cols_ were found during fitting.
+        if self.na_imputed_cols_:
+            for col in self.na_imputed_cols_:
+                X[col + "_missing"] = X[col].isnull().astype(int)
+
+        if not is_fitting and self.median_values_:
+            for col in self.median_values_:  # Iterate over keys if it's a dict
+                if col in X.columns:
+                    X[col] = X[col].fillna(self.median_values_[col])
+
+        return X
+
+    def _preprocess_X_fit(self, X, X_names, sample_weight):
+        if sample_weight.size > 0:
+            if sample_weight.ndim != 1:
+                raise ValueError("sample_weight must be a 1D array.")
+            if len(sample_weight) != X.shape[0]:
+                raise ValueError(
+                    "sample_weight must have the same number of rows as X."
+                )
+            if np.any(np.isnan(sample_weight)) or np.any(np.isinf(sample_weight)):
+                raise ValueError("sample_weight cannot contain nan or infinite values.")
+            if np.any(sample_weight < 0):
+                raise ValueError("sample_weight cannot contain negative values.")
+        X = self._common_X_preprocessing(X, is_fitting=True, X_names=X_names)
+        self.median_values_ = {}
+        numeric_cols_for_median = [col for col in X.columns if "_missing" not in col]
+        for col in numeric_cols_for_median:
+            missing_mask = X[col].isnull()
+
+            if sample_weight.size > 0:
+                valid_indices = ~missing_mask
+                col_data = X.loc[valid_indices, col]
+                col_weights = sample_weight[valid_indices]
+                if col_data.empty:
+                    median_val = 0
+                else:
+                    col_data_np = col_data.to_numpy()
+                    sort_indices = np.argsort(col_data_np, kind="stable")
+                    sorted_data = col_data_np[sort_indices]
+                    sorted_weights = col_weights[sort_indices]
+
+                    cumulative_weights = np.cumsum(sorted_weights)
+                    total_weight = cumulative_weights[-1]
+
+                    median_weight_index = np.searchsorted(
+                        cumulative_weights, total_weight / 2.0
+                    )
+                    if median_weight_index >= len(sorted_data):
+                        median_weight_index = len(sorted_data) - 1
+                    median_val = sorted_data[median_weight_index]
+            else:
+                median_val = X[col].median()
+
+            if pd.isna(median_val):
+                median_val = 0
+
+            self.median_values_[col] = median_val
+            X[col] = X[col].fillna(median_val)
+
+        self.final_training_columns_ = list(X.columns)
+        return X.values.astype(np.float64), list(X.columns)
+
+    def _preprocess_X_predict(self, X):
+        X = self._common_X_preprocessing(X, is_fitting=False)
+
+        # Enforce column order from training if it was set.
+        if self.final_training_columns_:
+            X = X[self.final_training_columns_]
+
+        return X.values.astype(np.float64)
+
+    def __setstate__(self, state):
+        """Handles unpickling for backward compatibility."""
+        self.__dict__.update(state)
+
+        # For backward compatibility, initialize new attributes to None if they don't exist,
+        # indicating the model was trained before these features were introduced.
+        new_attributes = [
+            "X_names_",
+            "categorical_features_",
+            "ohe_columns_",
+            "na_imputed_cols_",
+            "median_values_",
+            "final_training_columns_",
+        ]
+        for attr in new_attributes:
+            if not hasattr(self, attr):
+                setattr(self, attr, None)
+
+
+class APLRRegressor(BaseAPLR):
     def __init__(
         self,
         m: int = 3000,
@@ -127,6 +288,14 @@ class APLRRegressor:
         self.mean_bias_correction = mean_bias_correction
         self.faster_convergence = faster_convergence
 
+        # Data transformations
+        self.median_values_ = {}
+        self.categorical_features_ = []
+        self.ohe_columns_ = []
+        self.na_imputed_cols_ = []
+        self.X_names_ = []
+        self.final_training_columns_ = []
+
         # Creating aplr_cpp and setting parameters
         self.APLRRegressor = aplr_cpp.APLRRegressor()
         self.__set_params_cpp()
@@ -192,7 +361,7 @@ class APLRRegressor:
 
     def fit(
         self,
-        X: FloatMatrix,
+        X: Union[pd.DataFrame, FloatMatrix],
         y: FloatVector,
         sample_weight: FloatVector = np.empty(0),
         X_names: List[str] = [],
@@ -207,12 +376,16 @@ class APLRRegressor:
         predictor_penalties_for_interactions: List[float] = [],
         predictor_min_observations_in_split: List[int] = [],
     ):
+        self._validate_X_fit_rows(X)
         self.__set_params_cpp()
+        X_transformed, X_names_transformed = self._preprocess_X_fit(
+            X, X_names, sample_weight
+        )
         self.APLRRegressor.fit(
-            X,
+            X_transformed,
             y,
             sample_weight,
-            X_names,
+            X_names_transformed,
             cv_observations,
             prioritized_predictors_indexes,
             monotonic_constraints,
@@ -226,42 +399,65 @@ class APLRRegressor:
         )
 
     def predict(
-        self, X: FloatMatrix, cap_predictions_to_minmax_in_training: bool = True
+        self,
+        X: Union[pd.DataFrame, FloatMatrix],
+        cap_predictions_to_minmax_in_training: bool = True,
     ) -> FloatVector:
         if self.link_function == "custom_function":
             self.APLRRegressor.calculate_custom_transform_linear_predictor_to_predictions_function = (
                 self.calculate_custom_transform_linear_predictor_to_predictions_function
             )
-        return self.APLRRegressor.predict(X, cap_predictions_to_minmax_in_training)
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRRegressor.predict(
+            X_transformed, cap_predictions_to_minmax_in_training
+        )
 
     def set_term_names(self, X_names: List[str]):
         self.APLRRegressor.set_term_names(X_names)
 
     def calculate_feature_importance(
-        self, X: FloatMatrix, sample_weight: FloatVector = np.empty(0)
+        self,
+        X: Union[pd.DataFrame, FloatMatrix],
+        sample_weight: FloatVector = np.empty(0),
     ) -> FloatVector:
-        return self.APLRRegressor.calculate_feature_importance(X, sample_weight)
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRRegressor.calculate_feature_importance(
+            X_transformed, sample_weight
+        )
 
     def calculate_term_importance(
-        self, X: FloatMatrix, sample_weight: FloatVector = np.empty(0)
+        self,
+        X: Union[pd.DataFrame, FloatMatrix],
+        sample_weight: FloatVector = np.empty(0),
     ) -> FloatVector:
-        return self.APLRRegressor.calculate_term_importance(X, sample_weight)
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRRegressor.calculate_term_importance(
+            X_transformed, sample_weight
+        )
 
-    def calculate_local_feature_contribution(self, X: FloatMatrix) -> FloatMatrix:
-        return self.APLRRegressor.calculate_local_feature_contribution(X)
+    def calculate_local_feature_contribution(
+        self, X: Union[pd.DataFrame, FloatMatrix]
+    ) -> FloatMatrix:
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRRegressor.calculate_local_feature_contribution(X_transformed)
 
-    def calculate_local_term_contribution(self, X: FloatMatrix) -> FloatMatrix:
-        return self.APLRRegressor.calculate_local_term_contribution(X)
+    def calculate_local_term_contribution(
+        self, X: Union[pd.DataFrame, FloatMatrix]
+    ) -> FloatMatrix:
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRRegressor.calculate_local_term_contribution(X_transformed)
 
     def calculate_local_contribution_from_selected_terms(
-        self, X: FloatMatrix, predictor_indexes: List[int]
+        self, X: Union[pd.DataFrame, FloatMatrix], predictor_indexes: List[int]
     ) -> FloatVector:
+        X_transformed = self._preprocess_X_predict(X)
         return self.APLRRegressor.calculate_local_contribution_from_selected_terms(
-            X, predictor_indexes
+            X_transformed, predictor_indexes
         )
 
-    def calculate_terms(self, X: FloatMatrix) -> FloatMatrix:
-        return self.APLRRegressor.calculate_terms(X)
+    def calculate_terms(self, X: Union[pd.DataFrame, FloatMatrix]) -> FloatMatrix:
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRRegressor.calculate_terms(X_transformed)
 
     def get_term_names(self) -> List[str]:
         return self.APLRRegressor.get_term_names()
@@ -483,7 +679,7 @@ class APLRRegressor:
         return self
 
 
-class APLRClassifier:
+class APLRClassifier(BaseAPLR):
     def __init__(
         self,
         m: int = 3000,
@@ -534,6 +730,14 @@ class APLRClassifier:
         self.max_terms = max_terms
         self.ridge_penalty = ridge_penalty
 
+        # Data transformations
+        self.median_values_ = {}
+        self.categorical_features_ = []
+        self.ohe_columns_ = []
+        self.na_imputed_cols_ = []
+        self.X_names_ = []
+        self.final_training_columns_ = []
+
         # Creating aplr_cpp and setting parameters
         self.APLRClassifier = aplr_cpp.APLRClassifier()
         self.__set_params_cpp()
@@ -571,8 +775,8 @@ class APLRClassifier:
 
     def fit(
         self,
-        X: FloatMatrix,
-        y: List[str],
+        X: Union[pd.DataFrame, FloatMatrix],
+        y: Union[FloatVector, List[str]],
         sample_weight: FloatVector = np.empty(0),
         X_names: List[str] = [],
         cv_observations: IntMatrix = np.empty([0, 0]),
@@ -584,12 +788,22 @@ class APLRClassifier:
         predictor_penalties_for_interactions: List[float] = [],
         predictor_min_observations_in_split: List[int] = [],
     ):
+        self._validate_X_fit_rows(X)
         self.__set_params_cpp()
+        X_transformed, X_names_transformed = self._preprocess_X_fit(
+            X, X_names, sample_weight
+        )
+
+        if isinstance(y, np.ndarray):
+            y = y.astype(str).tolist()
+        elif isinstance(y, list) and y and not isinstance(y[0], str):
+            y = [str(val) for val in y]
+
         self.APLRClassifier.fit(
-            X,
+            X_transformed,
             y,
             sample_weight,
-            X_names,
+            X_names_transformed,
             cv_observations,
             prioritized_predictors_indexes,
             monotonic_constraints,
@@ -603,19 +817,30 @@ class APLRClassifier:
         self.classes_ = np.arange(len(self.APLRClassifier.get_categories()))
 
     def predict_class_probabilities(
-        self, X: FloatMatrix, cap_predictions_to_minmax_in_training: bool = False
+        self,
+        X: Union[pd.DataFrame, FloatMatrix],
+        cap_predictions_to_minmax_in_training: bool = False,
     ) -> FloatMatrix:
+        X_transformed = self._preprocess_X_predict(X)
         return self.APLRClassifier.predict_class_probabilities(
-            X, cap_predictions_to_minmax_in_training
+            X_transformed, cap_predictions_to_minmax_in_training
         )
 
     def predict(
-        self, X: FloatMatrix, cap_predictions_to_minmax_in_training: bool = False
+        self,
+        X: Union[pd.DataFrame, FloatMatrix],
+        cap_predictions_to_minmax_in_training: bool = False,
     ) -> List[str]:
-        return self.APLRClassifier.predict(X, cap_predictions_to_minmax_in_training)
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRClassifier.predict(
+            X_transformed, cap_predictions_to_minmax_in_training
+        )
 
-    def calculate_local_feature_contribution(self, X: FloatMatrix) -> FloatMatrix:
-        return self.APLRClassifier.calculate_local_feature_contribution(X)
+    def calculate_local_feature_contribution(
+        self, X: Union[pd.DataFrame, FloatMatrix]
+    ) -> FloatMatrix:
+        X_transformed = self._preprocess_X_predict(X)
+        return self.APLRClassifier.calculate_local_feature_contribution(X_transformed)
 
     def get_categories(self) -> List[str]:
         return self.APLRClassifier.get_categories()
@@ -724,7 +949,7 @@ class APLRTuner:
         grid = [dict(zip(keys, combination)) for combination in combinations]
         return grid
 
-    def fit(self, X: FloatMatrix, y: FloatVector, **kwargs):
+    def fit(self, X: Union[pd.DataFrame, FloatMatrix], y: FloatVector, **kwargs):
         self.cv_results: List[Dict[str, float]] = []
         best_validation_result = np.inf
         for params in self.parameter_grid:
@@ -742,10 +967,14 @@ class APLRTuner:
                 self.best_model = model
         self.cv_results = sorted(self.cv_results, key=lambda x: x["cv_error"])
 
-    def predict(self, X: FloatMatrix, **kwargs) -> Union[FloatVector, List[str]]:
+    def predict(
+        self, X: Union[pd.DataFrame, FloatMatrix], **kwargs
+    ) -> Union[FloatVector, List[str]]:
         return self.best_model.predict(X, **kwargs)
 
-    def predict_class_probabilities(self, X: FloatMatrix, **kwargs) -> FloatMatrix:
+    def predict_class_probabilities(
+        self, X: Union[pd.DataFrame, FloatMatrix], **kwargs
+    ) -> FloatMatrix:
         if self.is_regressor == False:
             return self.best_model.predict_class_probabilities(X, **kwargs)
         else:
@@ -753,7 +982,9 @@ class APLRTuner:
                 "predict_class_probabilities is only possible when is_regressor is False"
             )
 
-    def predict_proba(self, X: FloatMatrix, **kwargs) -> FloatMatrix:
+    def predict_proba(
+        self, X: Union[pd.DataFrame, FloatMatrix], **kwargs
+    ) -> FloatMatrix:
         return self.predict_class_probabilities(X, **kwargs)
 
     def get_best_estimator(self) -> Union[APLRClassifier, APLRRegressor]:

aplr-10.18.1.dist-info/METADATA

@@ -0,0 +1,34 @@
+Metadata-Version: 2.4
+Name: aplr
+Version: 10.18.1
+Summary: Automatic Piecewise Linear Regression
+Home-page: https://github.com/ottenbreit-data-science/aplr
+Author: Mathias von Ottenbreit
+Author-email: ottenbreitdatascience@gmail.com
+License: MIT
+Platform: Windows
+Platform: Linux
+Platform: MacOS
+Classifier: License :: OSI Approved :: MIT License
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.11
+Requires-Dist: pandas>=1.0.0
+Provides-Extra: plots
+Requires-Dist: matplotlib>=3.0; extra == "plots"
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: home-page
+Dynamic: license
+Dynamic: license-file
+Dynamic: platform
+Dynamic: provides-extra
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
+
+The documentation for Automatic Piecewise Linear Regression is available at [https://github.com/ottenbreit-data-science/aplr](https://github.com/ottenbreit-data-science/aplr).

aplr-10.18.1.dist-info/RECORD

@@ -0,0 +1,8 @@
+aplr_cpp.cp310-win32.pyd,sha256=rmLQxuwwX35W_v6uEoYl4PnnwAre5K4-VCIkVkmW1TQ,596992
+aplr/__init__.py,sha256=oDFSgVytP_qQ8ilun6oHxKr-DYEeqjEQp5FciX45lls,21
+aplr/aplr.py,sha256=YOXF2a5CpC6kYoLGaDOaYIXMSRrWtSwvrbxpjYzPyA0,42653
+aplr-10.18.1.dist-info/licenses/LICENSE,sha256=YOMo-RaL4P7edMZGD96-NskKpxyMZdP3-WiiMMmihNk,1134
+aplr-10.18.1.dist-info/METADATA,sha256=lj9HjkVuCgcD7cKxWOpRMXrocW4r-Zzpa4HQm3UkdkA,1048
+aplr-10.18.1.dist-info/WHEEL,sha256=GWZF0cboiU4MhsG0baPl8rrtCaXFSLW25384gp3vddM,97
+aplr-10.18.1.dist-info/top_level.txt,sha256=DXVC0RIFGpzVnPeKWAZTXQdJheOEZL51Wip6Fx7zbR4,14
+aplr-10.18.1.dist-info/RECORD,,

aplr-10.17.1.dist-info/METADATA

@@ -1,68 +0,0 @@
-Metadata-Version: 2.4
-Name: aplr
-Version: 10.17.1
-Summary: Automatic Piecewise Linear Regression
-Home-page: https://github.com/ottenbreit-data-science/aplr
-Author: Mathias von Ottenbreit
-Author-email: ottenbreitdatascience@gmail.com
-License: MIT
-Platform: Windows
-Platform: Linux
-Platform: MacOS
-Classifier: License :: OSI Approved :: MIT License
-Requires-Python: >=3.8
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: numpy>=1.11
-Provides-Extra: plots
-Requires-Dist: matplotlib>=3.0; extra == "plots"
-Dynamic: author
-Dynamic: author-email
-Dynamic: classifier
-Dynamic: description
-Dynamic: description-content-type
-Dynamic: home-page
-Dynamic: license
-Dynamic: license-file
-Dynamic: platform
-Dynamic: provides-extra
-Dynamic: requires-dist
-Dynamic: requires-python
-Dynamic: summary
-
-# APLR
-**Automatic Piecewise Linear Regression**
-
-## About
-APLR allows you to build predictive and interpretable regression or classification machine learning models in Python, using the Automatic Piecewise Linear Regression (APLR) methodology developed by Mathias von Ottenbreit. APLR often rivals tree-based methods in predictive accuracy, while offering smoother, more interpretable predictions. 
-
-For further details, see the [documentation](https://github.com/ottenbreit-data-science/aplr/tree/main/documentation). You may also read the published article for additional insights: [Link 1](https://link.springer.com/article/10.1007/s00180-024-01475-4) and [Link 2](https://rdcu.be/dz7bF). Additional functionality has been added since the article was published.
-
-## Installation
-To install APLR, use the following command:
-
-```bash
-pip install aplr
-```
-
-To include dependencies for plotting, use this command instead:
-
-```bash
-pip install aplr[plots]
-```
-
-## Availability
-APLR is available for Windows, most Linux distributions, and macOS.
-
-## Usage
-Example Python scripts are available [here](https://github.com/ottenbreit-data-science/aplr/tree/main/examples).
-
-## Sponsorship
-Consider sponsoring Von Ottenbreit Data Science by clicking the **Sponsor** button on the repository. Sufficient funding will help maintain and further develop APLR.
-
-## API Reference
-- [API reference for regression](https://github.com/ottenbreit-data-science/aplr/blob/main/API_REFERENCE_FOR_REGRESSION.md)
-- [API reference for classification](https://github.com/ottenbreit-data-science/aplr/blob/main/API_REFERENCE_FOR_CLASSIFICATION.md)
-
-## Contact Information
-For inquiries, please email: [ottenbreitdatascience@gmail.com](mailto:ottenbreitdatascience@gmail.com)

aplr-10.17.1.dist-info/RECORD

@@ -1,8 +0,0 @@
-aplr_cpp.cp310-win32.pyd,sha256=c_nkUTh9KaMBtfBuC22-JDGfEx3LMpvPd9qUr2l0ZaY,596480
-aplr/__init__.py,sha256=oDFSgVytP_qQ8ilun6oHxKr-DYEeqjEQp5FciX45lls,21
-aplr/aplr.py,sha256=obmpNNppOrnzQxhYzShYzNe02lQv17-YRWUklf8IXds,33302
-aplr-10.17.1.dist-info/licenses/LICENSE,sha256=YOMo-RaL4P7edMZGD96-NskKpxyMZdP3-WiiMMmihNk,1134
-aplr-10.17.1.dist-info/METADATA,sha256=crEeFMmcUldMHf6eZnAi2zAbnQIWcQBOLKZeVZdej58,2627
-aplr-10.17.1.dist-info/WHEEL,sha256=GWZF0cboiU4MhsG0baPl8rrtCaXFSLW25384gp3vddM,97
-aplr-10.17.1.dist-info/top_level.txt,sha256=DXVC0RIFGpzVnPeKWAZTXQdJheOEZL51Wip6Fx7zbR4,14
-aplr-10.17.1.dist-info/RECORD,,