refactor: initializers module

docs/source/api/initializers/index.rst

@@ -40,5 +40,6 @@ Enumeration types that define available strategies for initialization.
    :nosignatures:
    :template: class.rst
 
-   ClusterMatchStrategy
+   MatchingMethod
+   ScoringMethod
    EstimationStrategy

rework_pysatl_mpest/initializers/__init__.py

@@ -9,64 +9,24 @@
 
 initializers provide good starting points for EM algorithm and other optimization
 methods, helping to avoid poor local optima and improving convergence.
-
-**Usage Example**
-
-.. code-block:: python
-
-    >>> from rework_pysatl_mpest import Exponential
-    >>> import numpy as np
-    >>> from sklearn.cluster import KMeans
-    >>> from rework_pysatl_mpest.initializers import ClusterizeInitializer
-    >>> from rework_pysatl_mpest.initializers import ClusterMatchStrategy, EstimationStrategy
-
-    >>> # Create initializer with KMeans clustering
-    >>> initializer_cluster = ClusterizeInitializer(
-    ...     is_accurate=True,
-    ...     is_soft=False,
-    ...     clusterizer=KMeans(n_clusters=3)
-    ... )
-
-    >>> # Create distribution models to initialize
-    >>> distributions = [Exponential(loc=0.0, rate=0.1),
-    >>>Exponential(loc=5.0, rate=0.05), Exponential(loc=10.0, rate=0.01)]
-
-    >>> # Generate sample data
-    >>> X = np.linspace(0.01, 25.0, 300)
-
-    >>> # Perform initialization
-    >>> mixture_model = initializer_cluster.perform(
-    ...     X=X,
-    ...     dists=distributions,
-    ...     cluster_match_strategy=ClusterMatchStrategy.AKAIKE,
-    ...     estimation_strategies=[EstimationStrategy.QFUNCTION] * len(distributions)
-    ... )
-
-    >>> # The mixture model is now initialized with estimated parameters
-    >>> print(f"Number of components: {len(mixture_model.components)}")
-    >>> print(f"Weights: {mixture_model.weights}")
 """
 
 __author__ = "Viktor Khanukaev"
 __copyright__ = "Copyright (c) 2025 PySATL project"
 __license__ = "SPDX-License-Identifier: MIT"
 
-from ._estimation_strategies.q_function import q_function_strategy, q_function_strategy_exponential
-from .cluster_match_strategy import (
-    match_clusters_for_models_akaike,
-    match_clusters_for_models_log_likelihood,
+from .clusterize import (
+    ClusterizeInitializer,
+    EstimationStrategy,
+    MatchingMethod,
+    ScoringMethod,
 )
-from .clusterize_initializer import ClusterizeInitializer
 from .initializer import Initializer
-from .strategies import ClusterMatchStrategy, EstimationStrategy
 
 __all__ = [
-    "ClusterMatchStrategy",
     "ClusterizeInitializer",
     "EstimationStrategy",
     "Initializer",
-    "match_clusters_for_models_akaike",
-    "match_clusters_for_models_log_likelihood",
-    "q_function_strategy",
-    "q_function_strategy_exponential",
+    "MatchingMethod",
+    "ScoringMethod",
 ]

rework_pysatl_mpest/initializers/_estimation_strategies/__init__.py

@@ -0,0 +1,14 @@
+"""
+A subpackage containing estimation strategies for initialization.
+"""
+
+__author__ = "Viktor Khanukaev"
+__copyright__ = "Copyright (c) 2025 PySATL project"
+__license__ = "SPDX-License-Identifier: MIT"
+
+from .q_function import q_function_strategy, q_function_strategy_exponential
+
+__all__ = [
+    "q_function_strategy",
+    "q_function_strategy_exponential",
+]

rework_pysatl_mpest/initializers/_estimation_strategies/q_function.py

@@ -11,11 +11,10 @@
 
 import numpy as np
 
-from rework_pysatl_mpest.distributions.continuous_dist import ContinuousDistribution
-from rework_pysatl_mpest.distributions.exponential import Exponential
-from rework_pysatl_mpest.optimizers import Optimizer
+from ...distributions import ContinuousDistribution, Exponential
+from ...optimizers import Optimizer
 
-NUMERICAL_TOLERANCE = 0.33
+NUMERICAL_TOLERANCE = 1e-6
 
 
 @singledispatch
@@ -164,7 +163,7 @@ def q_function_strategy_exponential(
     >>> print(f"Estimated loc: {params['loc']:.3f}, rate: {params['rate']:.3f}")
     """
 
-    new_params = {}
+    new_params: dict = {}
     N_j = np.sum(H_j).item()
 
     if np.any(H_j > NUMERICAL_TOLERANCE):