aek-auto-mlbuilder 0.3.2__tar.gz → 0.5.0__tar.gz

{aek_auto_mlbuilder-0.3.2 → aek_auto_mlbuilder-0.5.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aek-auto-mlbuilder
-Version: 0.3.2
+Version: 0.5.0
 Summary: Automatic ML model builder in Python
 Home-page: https://github.com/alpemre8/aek-auto-mlbuilder
 Author: Alp Emre Karaahmet

{aek_auto_mlbuilder-0.3.2 → aek_auto_mlbuilder-0.5.0}/aek_auto_mlbuilder/__init__.py

@@ -2,4 +2,6 @@ from .base import BaseModel
 from .utils import split_data
 from .linear_regression import LinearRegressor
 from .logistic_regression import LogisticClassifier
-from .decision_tree import DecisionTreeModel
+from .decision_tree import DecisionTreeModel
+from .knn import KNNModel
+from .random_forest import RandomForestModel

aek_auto_mlbuilder-0.5.0/aek_auto_mlbuilder/knn.py

@@ -0,0 +1,48 @@
+from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor
+from sklearn.preprocessing import StandardScaler
+from sklearn.pipeline import make_pipeline
+from .base import BaseModel
+
+
+class KNNModel(BaseModel):
+    """
+    KNN model for classification for classification or regression
+    Use "task" for "classification" or "regression"
+    Brute-force search is using
+    """
+    def __init__(self, task="classification", param_grid=None):
+        super().__init__()
+        self.task = task
+        self.param_grid = param_grid or {
+            "n_neighbors": [3, 5, 7, 9],
+            "weights": ["uniform", "distance"],
+            "p": [1, 2] #1: manhattan, 2: euclidean
+        }
+
+    def train(self, X, y):
+        best_score = -float("inf")
+        best_model = None
+
+        if self.task.lower() == "classification":
+            ModelClass = KNeighborsClassifier
+        elif self.task.lower() == "regression":
+            ModelClass = KNeighborsRegressor
+        else:
+            raise ValueError("task must be 'classification' or 'regression'")
+        
+        for n in self.param_grid["n_neighbors"]:
+            for weights in self.param_grid["weights"]:
+                for p in self.param_grid["p"]:
+                    model = make_pipeline(
+                        StandardScaler(),
+                        ModelClass(n_neighbors=n, weights=weights, p=p)
+                    )
+                    model.fit(X, y)
+                    score = model.score(X, y)
+                    if score > best_score:
+                        best_score = score
+                        best_model = model
+
+        self.best_model = best_model
+        self.best_score = best_score
+        return self.best_model

aek_auto_mlbuilder-0.5.0/aek_auto_mlbuilder/random_forest.py

@@ -0,0 +1,50 @@
+from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
+from .base import BaseModel
+
+class RandomForestModel(BaseModel):
+    """
+    Random Forest model for classification or regression
+    Use task either "classification" or "regression"
+    brute force method is using
+    """
+    def __init__(self, task="classification", param_grid=None):
+        super().__init__()
+        self.task = task
+        self.param_grid = param_grid or {
+            "n_estimators": [50, 100, 200],
+            "max_depth": [None, 5, 10],
+            "min_samples_split": [2, 5],
+            "min_samples_leaf": [1, 2]
+        }
+    def train(self, X, y):
+        best_score = -float("inf")
+        best_model = None
+
+        if self.task.lower() == "classification":
+            ModelClass = RandomForestClassifier
+        elif self.task.lower() == "regression":
+            ModelClass = RandomForestRegressor
+        else:
+            raise ValueError("task must be 'classification' or 'regression'")
+        
+        for n_estimators in self.param_grid["n_estimators"]:
+            for max_depth in self.param_grid["max_depth"]:
+                for min_samples_split in self.param_grid["min_samples_split"]:
+                    for min_samples_leaf in self.param_grid["min_samples_leaf"]:
+                        model = ModelClass(
+                            n_estimators=n_estimators,
+                            max_depth=max_depth,
+                            min_samples_split=min_samples_split,
+                            min_samples_leaf=min_samples_leaf,
+                            random_state=42,
+                            n_jobs=-1
+                        )
+                        model.fit(X, y)
+                        score = model.score(X, y)
+                        if score > best_score:
+                            best_score = score
+                            best_model = model
+        self.best_model = best_model
+        self.best_score = best_score
+        return self.best_model
+    

{aek_auto_mlbuilder-0.3.2 → aek_auto_mlbuilder-0.5.0}/aek_auto_mlbuilder.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aek-auto-mlbuilder
-Version: 0.3.2
+Version: 0.5.0
 Summary: Automatic ML model builder in Python
 Home-page: https://github.com/alpemre8/aek-auto-mlbuilder
 Author: Alp Emre Karaahmet

{aek_auto_mlbuilder-0.3.2 → aek_auto_mlbuilder-0.5.0}/aek_auto_mlbuilder.egg-info/SOURCES.txt

@@ -5,8 +5,10 @@ setup.py
 aek_auto_mlbuilder/__init__.py
 aek_auto_mlbuilder/base.py
 aek_auto_mlbuilder/decision_tree.py
+aek_auto_mlbuilder/knn.py
 aek_auto_mlbuilder/linear_regression.py
 aek_auto_mlbuilder/logistic_regression.py
+aek_auto_mlbuilder/random_forest.py
 aek_auto_mlbuilder/utils.py
 aek_auto_mlbuilder.egg-info/PKG-INFO
 aek_auto_mlbuilder.egg-info/SOURCES.txt

{aek_auto_mlbuilder-0.3.2 → aek_auto_mlbuilder-0.5.0}/setup.py

@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
 
 setup(
     name="aek-auto-mlbuilder",        # PyPI dağıtım adı
-    version="0.3.2",
+    version="0.5.0",
     description="Automatic ML model builder in Python",
     long_description=open("README.md", encoding="utf-8").read(),
     long_description_content_type="text/markdown",