feat: return new model when calling fit

docs/tutorials/machine_learning.ipynb

@@ -60,7 +60,7 @@
     "from safeds.ml.regression import LinearRegression\n",
     "\n",
     "model = LinearRegression()\n",
-    "model.fit(tagged_table)"
+    "fitted_model = model.fit(tagged_table)"
    ],
    "metadata": {
     "collapsed": false
@@ -71,7 +71,7 @@
    "source": [
     "## Predicting new values\n",
     "\n",
-    "The `fit` method trains the model in place. This means that the model object is modified and can be used to make predictions. Predictions are made by calling the `predict` method on the model object. The `predict` method takes a `Table` as input and returns a `Table` with the predictions:"
+    "The `fit` method returns the fitted model, the original model is **not** changed. Predictions are made by calling the `predict` method on the fitted model. The `predict` method takes a `Table` as input and returns a `Table` with the predictions:"
    ],
    "metadata": {
     "collapsed": false
@@ -87,7 +87,7 @@
     "    \"b\": [2, 0, 5, 2, 7],\n",
     "    \"c\": [1, 4, 3, 2, 1]})\n",
     "\n",
-    "model.predict(dataset=test_set)\n"
+    "fitted_model.predict(dataset=test_set)\n"
    ],
    "metadata": {
     "collapsed": false

src/safeds/ml/_util_sklearn.py

@@ -1,4 +1,4 @@
-from typing import Any
+from typing import Any, Optional
 
 from safeds.data.tabular.containers import Table, TaggedTable
 from safeds.exceptions import LearningError, PredictionError
@@ -34,7 +34,7 @@ def fit(model: Any, tagged_table: TaggedTable) -> None:
 
 
 # noinspection PyProtectedMember
-def predict(model: Any, dataset: Table, target_name: str) -> TaggedTable:
+def predict(model: Any, dataset: Table, target_name: Optional[str]) -> TaggedTable:
     """
     Predict a target vector using a dataset containing feature vectors. The model has to be trained first.
 
@@ -57,6 +57,10 @@ def predict(model: Any, dataset: Table, target_name: str) -> TaggedTable:
     PredictionError
         If predicting with the given dataset failed.
     """
+
+    if model is None or target_name is None:
+        raise PredictionError("The model was not trained")
+
     dataset_df = dataset._data
     dataset_df.columns = dataset.schema.get_column_names()
     try:

src/safeds/ml/classification/_ada_boost.py

@@ -1,3 +1,7 @@
+from __future__ import annotations
+
+from typing import Optional
+
 from safeds.data.tabular.containers import Table, TaggedTable
 from safeds.ml._util_sklearn import fit, predict
 from sklearn.ensemble import AdaBoostClassifier as sk_AdaBoostClassifier
@@ -12,25 +16,38 @@ class AdaBoost(Classifier):
     """
 
     def __init__(self) -> None:
-        self._wrapped_classifier = sk_AdaBoostClassifier()
-        self._target_name = ""
+        self._wrapped_classifier: Optional[sk_AdaBoostClassifier] = None
+        self._target_name: Optional[str] = None
 
-    def fit(self, training_set: TaggedTable) -> None:
+    def fit(self, training_set: TaggedTable) -> AdaBoost:
         """
-        Fit this model given a tagged table.
+        Create a new classifier based on this one and fit it with the given training data. This classifier is not
+        modified.
 
         Parameters
         ----------
         training_set : TaggedTable
-            The tagged table containing the feature and target vectors.
+            The training data containing the feature and target vectors.
+
+        Returns
+        -------
+        fitted_classifier : AdaBoost
+            The fitted classifier.
 
         Raises
         ------
         LearningError
-            If the tagged table contains invalid values or if the training failed.
+            If the training data contains invalid values or if the training failed.
         """
-        fit(self._wrapped_classifier, training_set)
-        self._target_name = training_set.target.name
+
+        wrapped_classifier = sk_AdaBoostClassifier()
+        fit(wrapped_classifier, training_set)
+
+        result = AdaBoost()
+        result._wrapped_classifier = wrapped_classifier
+        result._target_name = training_set.target.name
+
+        return result
 
     def predict(self, dataset: Table) -> TaggedTable:
         """

src/safeds/ml/classification/_classifier.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from abc import ABC, abstractmethod
 
 from safeds.data.tabular.containers import Table, TaggedTable
@@ -10,19 +12,25 @@ class Classifier(ABC):
     """
 
     @abstractmethod
-    def fit(self, training_set: TaggedTable) -> None:
+    def fit(self, training_set: TaggedTable) -> Classifier:
         """
-        Fit this model given a tagged table.
+        Create a new classifier based on this one and fit it with the given training data. This classifier is not
+        modified.
 
         Parameters
         ----------
         training_set : TaggedTable
-            The tagged table containing the feature and target vectors.
+            The training data containing the feature and target vectors.
+
+        Returns
+        -------
+        fitted_classifier : Classifier
+            The fitted classifier.
 
         Raises
         ------
         LearningError
-            If the tagged table contains invalid values or if the training failed.
+            If the training data contains invalid values or if the training failed.
         """
 
     @abstractmethod

src/safeds/ml/classification/_decision_tree.py

@@ -1,3 +1,7 @@
+from __future__ import annotations
+
+from typing import Optional
+
 from safeds.data.tabular.containers import Table, TaggedTable
 from safeds.ml._util_sklearn import fit, predict
 from sklearn.tree import DecisionTreeClassifier as sk_DecisionTreeClassifier
@@ -12,25 +16,38 @@ class DecisionTree(Classifier):
     """
 
     def __init__(self) -> None:
-        self._wrapped_classifier = sk_DecisionTreeClassifier()
-        self._target_name = ""
+        self._wrapped_classifier: Optional[sk_DecisionTreeClassifier] = None
+        self._target_name: Optional[str] = None
 
-    def fit(self, training_set: TaggedTable) -> None:
+    def fit(self, training_set: TaggedTable) -> DecisionTree:
         """
-        Fit this model given a tagged table.
+        Create a new classifier based on this one and fit it with the given training data. This classifier is not
+        modified.
 
         Parameters
         ----------
         training_set : TaggedTable
-            The tagged table containing the feature and target vectors.
+            The training data containing the feature and target vectors.
+
+        Returns
+        -------
+        fitted_classifier : DecisionTree
+            The fitted classifier.
 
         Raises
         ------
         LearningError
-            If the tagged table contains invalid values or if the training failed.
+            If the training data contains invalid values or if the training failed.
         """
-        fit(self._wrapped_classifier, training_set)
-        self._target_name = training_set.target.name
+
+        wrapped_classifier = sk_DecisionTreeClassifier()
+        fit(wrapped_classifier, training_set)
+
+        result = DecisionTree()
+        result._wrapped_classifier = wrapped_classifier
+        result._target_name = training_set.target.name
+
+        return result
 
     def predict(self, dataset: Table) -> TaggedTable:
         """

src/safeds/ml/classification/_gradient_boosting_classification.py

@@ -1,3 +1,7 @@
+from __future__ import annotations
+
+from typing import Optional
+
 from safeds.data.tabular.containers import Table, TaggedTable
 from safeds.ml._util_sklearn import fit, predict
 from sklearn.ensemble import GradientBoostingClassifier as sk_GradientBoostingClassifier
@@ -12,25 +16,38 @@ class GradientBoosting(Classifier):
     """
 
     def __init__(self) -> None:
-        self._wrapped_classifier = sk_GradientBoostingClassifier()
-        self._target_name = ""
+        self._wrapped_classifier: Optional[sk_GradientBoostingClassifier] = None
+        self._target_name: Optional[str] = None
 
-    def fit(self, training_set: TaggedTable) -> None:
+    def fit(self, training_set: TaggedTable) -> GradientBoosting:
         """
-        Fit this model given a tagged table.
+        Create a new classifier based on this one and fit it with the given training data. This classifier is not
+        modified.
 
         Parameters
         ----------
         training_set : TaggedTable
-            The tagged table containing the feature and target vectors.
+            The training data containing the feature and target vectors.
+
+        Returns
+        -------
+        fitted_classifier : GradientBoosting
+            The fitted classifier.
 
         Raises
         ------
         LearningError
-            If the tagged table contains invalid values or if the training failed.
+            If the training data contains invalid values or if the training failed.
         """
-        fit(self._wrapped_classifier, training_set)
-        self._target_name = training_set.target.name
+
+        wrapped_classifier = sk_GradientBoostingClassifier()
+        fit(wrapped_classifier, training_set)
+
+        result = GradientBoosting()
+        result._wrapped_classifier = wrapped_classifier
+        result._target_name = training_set.target.name
+
+        return result
 
     # noinspection PyProtectedMember
     def predict(self, dataset: Table) -> TaggedTable:

src/safeds/ml/classification/_k_nearest_neighbors.py

@@ -1,3 +1,7 @@
+from __future__ import annotations
+
+from typing import Optional
+
 from safeds.data.tabular.containers import Table, TaggedTable
 from safeds.ml._util_sklearn import fit, predict
 from sklearn.neighbors import KNeighborsClassifier as sk_KNeighborsClassifier
@@ -16,27 +20,39 @@ class KNearestNeighbors(Classifier):
     """
 
     def __init__(self, n_neighbors: int) -> None:
-        self._wrapped_classifier = sk_KNeighborsClassifier(
-            n_jobs=-1, n_neighbors=n_neighbors
-        )
-        self._target_name = ""
+        self._n_neighbors = n_neighbors
 
-    def fit(self, training_set: TaggedTable) -> None:
+        self._wrapped_classifier: Optional[sk_KNeighborsClassifier] = None
+        self._target_name: Optional[str] = None
+
+    def fit(self, training_set: TaggedTable) -> KNearestNeighbors:
         """
-        Fit this model given a tagged table.
+        Create a new classifier based on this one and fit it with the given training data. This classifier is not
+        modified.
 
         Parameters
         ----------
         training_set : TaggedTable
-            The tagged table containing the feature and target vectors.
+            The training data containing the feature and target vectors.
+
+        Returns
+        -------
+        fitted_classifier : KNearestNeighbors
+            The fitted classifier.
 
         Raises
         ------
         LearningError
-            If the tagged table contains invalid values or if the training failed.
+            If the training data contains invalid values or if the training failed.
         """
-        fit(self._wrapped_classifier, training_set)
-        self._target_name = training_set.target.name
+        wrapped_classifier = sk_KNeighborsClassifier(self._n_neighbors, n_jobs=-1)
+        fit(wrapped_classifier, training_set)
+
+        result = KNearestNeighbors(self._n_neighbors)
+        result._wrapped_classifier = wrapped_classifier
+        result._target_name = training_set.target.name
+
+        return result
 
     def predict(self, dataset: Table) -> TaggedTable:
         """

src/safeds/ml/classification/_logistic_regression.py

@@ -1,3 +1,7 @@
+from __future__ import annotations
+
+from typing import Optional
+
 from safeds.data.tabular.containers import Table, TaggedTable
 from safeds.ml._util_sklearn import fit, predict
 from sklearn.linear_model import LogisticRegression as sk_LogisticRegression
@@ -12,25 +16,38 @@ class LogisticRegression(Classifier):
     """
 
     def __init__(self) -> None:
-        self._wrapped_classifier = sk_LogisticRegression(n_jobs=-1)
-        self._target_name = ""
+        self._wrapped_classifier: Optional[sk_LogisticRegression] = None
+        self._target_name: Optional[str] = None
 
-    def fit(self, training_set: TaggedTable) -> None:
+    def fit(self, training_set: TaggedTable) -> LogisticRegression:
         """
-        Fit this model given a tagged table.
+        Create a new classifier based on this one and fit it with the given training data. This classifier is not
+        modified.
 
         Parameters
         ----------
         training_set : TaggedTable
-            The tagged table containing the feature and target vectors.
+            The training data containing the feature and target vectors.
+
+        Returns
+        -------
+        fitted_classifier : LogisticRegression
+            The fitted classifier.
 
         Raises
         ------
         LearningError
-            If the tagged table contains invalid values or if the training failed.
+            If the training data contains invalid values or if the training failed.
         """
-        fit(self._wrapped_classifier, training_set)
-        self._target_name = training_set.target.name
+
+        wrapped_classifier = sk_LogisticRegression(n_jobs=-1)
+        fit(wrapped_classifier, training_set)
+
+        result = LogisticRegression()
+        result._wrapped_classifier = wrapped_classifier
+        result._target_name = training_set.target.name
+
+        return result
 
     def predict(self, dataset: Table) -> TaggedTable:
         """