Merge branch 'develop'

adaa-polsl · Feb 27, 2024 · 19d97cb · 19d97cb
2 parents 236835a + 90780cd
commit 19d97cb
Show file tree

Hide file tree

Showing 6 changed files with 54 additions and 55 deletions.
diff --git a/rulekit/__init__.py b/rulekit/__init__.py
@@ -1,5 +1,5 @@
 # pylint: disable=missing-module-docstring
 from .main import RuleKit
 
-__VERSION__ = '1.7.5'
-__RULEKIT_RELEASE_VERSION__ = '1.7.4'
+__VERSION__ = '1.7.6'
+__RULEKIT_RELEASE_VERSION__ = '1.7.5'
diff --git a/rulekit/classification.py b/rulekit/classification.py
@@ -73,7 +73,7 @@ class RuleClassifier(BaseOperator, BaseClassifier):
 
     def __init__(
         self,
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         induction_measure: Measures = DEFAULT_PARAMS_VALUE['induction_measure'],
         pruning_measure: Union[Measures,
                                str] = DEFAULT_PARAMS_VALUE['pruning_measure'],
@@ -88,14 +88,14 @@ def __init__(
         max_rule_count: int = DEFAULT_PARAMS_VALUE['max_rule_count'],
         approximate_induction: bool = DEFAULT_PARAMS_VALUE['approximate_induction'],
         approximate_bins_count: int = DEFAULT_PARAMS_VALUE['approximate_bins_count'],
-        min_rule_covered: Optional[int] = None,
+        min_rule_covered: Optional[float] = None,
     ):
         """
         Parameters
         ----------
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         induction_measure : :class:`rulekit.params.Measures` = :class:`rulekit.params.\
             Measures.Correlation`
             measure used during induction; default measure is correlation
@@ -137,7 +137,7 @@ def __init__(
             data sets, results may change in future;
         approximate_bins_count: int = 100
             maximum number of bins for an attribute evaluated in the approximate induction.
-        min_rule_covered : int = None
+        min_rule_covered : float = None
             alias to `minsupp_new`. Parameter is deprecated and will be removed in the next major
             version, use `minsupp_new`
 
@@ -327,7 +327,7 @@ class ExpertRuleClassifier(ExpertKnowledgeOperator, RuleClassifier):
 
     def __init__(
         self,
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         induction_measure: Measures = DEFAULT_PARAMS_VALUE['induction_measure'],
         pruning_measure: Union[Measures,
                                str] = DEFAULT_PARAMS_VALUE['pruning_measure'],
@@ -352,14 +352,14 @@ def __init__(
             'preferred_conditions_per_rule'],
         preferred_attributes_per_rule: int = DEFAULT_PARAMS_VALUE[
             'preferred_attributes_per_rule'],
-        min_rule_covered: Optional[int] = None
+        min_rule_covered: Optional[float] = None
     ):
         """
         Parameters
         ----------
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         induction_measure : :class:`rulekit.params.Measures` = \
             :class:`rulekit.params.Measures.Correlation`
             measure used during induction; default measure is correlation
@@ -421,7 +421,7 @@ def __init__(
             maximum number of preferred conditions per rule; default: unlimited,
         preferred_attributes_per_rule : int = None
             maximum number of preferred attributes per rule; default: unlimited.
-        min_rule_covered : int = None
+        min_rule_covered : float = None
             alias to `minsupp_new`. Parameter is deprecated and will be removed in the next major
             version, use `minsupp_new`
 
@@ -555,7 +555,7 @@ def __init__(
         penalty_strength: float = DEFAULT_PARAMS_VALUE['penalty_strength'],
         penalty_saturation: float = DEFAULT_PARAMS_VALUE['penalty_saturation'],
 
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         induction_measure: Measures = DEFAULT_PARAMS_VALUE['induction_measure'],
         pruning_measure: Union[Measures,
                                str] = DEFAULT_PARAMS_VALUE['pruning_measure'],
@@ -585,9 +585,9 @@ def __init__(
             (s) - penalty strength; Default is 0.5
         penalty_saturation: float
             the value of p_new / P at which penalty reward saturates; Default is 0.2.
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be 
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         induction_measure : :class:`rulekit.params.Measures` = \
             :class:`rulekit.params.Measures.Correlation`
             measure used during induction; default measure is correlation

diff --git a/rulekit/params.py b/rulekit/params.py
@@ -59,8 +59,8 @@ class Measures(Enum):
 
 
 DEFAULT_PARAMS_VALUE = {
-    'minsupp_new': 5,
-    'min_rule_covered': 5,
+    'minsupp_new': 5.0,
+    'min_rule_covered': 5.0,
     'induction_measure': Measures.Correlation,
     'pruning_measure':  Measures.Correlation,
     'voting_measure': Measures.Correlation,
@@ -96,8 +96,8 @@ class Measures(Enum):
 class ModelsParams(BaseModel):
     """Model for validating models hyperparameters
     """
-    min_rule_covered: Optional[int] = None
-    minsupp_new: Optional[int] = DEFAULT_PARAMS_VALUE['minsupp_new']
+    min_rule_covered: Optional[float] = None
+    minsupp_new: Optional[float] = DEFAULT_PARAMS_VALUE['minsupp_new']
     induction_measure: Optional[Measures] = DEFAULT_PARAMS_VALUE['induction_measure']
     pruning_measure: Optional[Measures] = DEFAULT_PARAMS_VALUE['pruning_measure']
     voting_measure: Optional[Measures] = DEFAULT_PARAMS_VALUE['voting_measure']

diff --git a/rulekit/regression.py b/rulekit/regression.py
@@ -27,7 +27,7 @@ class RuleRegressor(BaseOperator):
 
     def __init__(
         self,
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         induction_measure: Measures = DEFAULT_PARAMS_VALUE['induction_measure'],
         pruning_measure: Union[Measures,
                                str] = DEFAULT_PARAMS_VALUE['pruning_measure'],
@@ -40,14 +40,14 @@ def __init__(
         complementary_conditions: bool = DEFAULT_PARAMS_VALUE['complementary_conditions'],
         mean_based_regression: bool = DEFAULT_PARAMS_VALUE['mean_based_regression'],
         max_rule_count: int = DEFAULT_PARAMS_VALUE['max_rule_count'],
-        min_rule_covered: Optional[int] = None,
+        min_rule_covered: Optional[float] = None,
     ):
         """
         Parameters
         ----------
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         induction_measure : :class:`rulekit.params.Measures` = \
             :class:`rulekit.params.Measures.Correlation`
             measure used during induction; default measure is correlation
@@ -82,7 +82,7 @@ def __init__(
         max_rule_count : int = 0
             Maximum number of rules to be generated (for classification data sets it applies 
             to a single class); 0 indicates no limit.
-        min_rule_covered : int = None
+        min_rule_covered : float = None
             alias to `minsupp_new`. Parameter is deprecated and will be removed in the next major
             version, use `minsupp_new`
 
@@ -176,7 +176,7 @@ class ExpertRuleRegressor(ExpertKnowledgeOperator, RuleRegressor):
 
     def __init__(
         self,
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         induction_measure: Measures = DEFAULT_PARAMS_VALUE['induction_measure'],
         pruning_measure: Union[Measures,
                                str] = DEFAULT_PARAMS_VALUE['pruning_measure'],
@@ -199,14 +199,14 @@ def __init__(
         preferred_attributes_per_rule: int = DEFAULT_PARAMS_VALUE[
             'preferred_attributes_per_rule'],
 
-        min_rule_covered: Optional[int] = None
+        min_rule_covered: Optional[float] = None
     ):
         """
         Parameters
         ----------
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         induction_measure : :class:`rulekit.params.Measures` = \
             :class:`rulekit.params.Measures.Correlation`
             measure used during induction; default measure is correlation
@@ -364,7 +364,7 @@ def __init__(
         penalty_strength: float = DEFAULT_PARAMS_VALUE['penalty_strength'],
         penalty_saturation: float = DEFAULT_PARAMS_VALUE['penalty_saturation'],
 
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         induction_measure: Measures = DEFAULT_PARAMS_VALUE['induction_measure'],
         pruning_measure: Union[Measures,
                                str] = DEFAULT_PARAMS_VALUE['pruning_measure'],
@@ -392,9 +392,9 @@ def __init__(
             (s) - penalty strength; Default is 0.5
         penalty_saturation: float
             the value of p_new / P at which penalty reward saturates; Default is 0.2.
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be 
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         induction_measure : :class:`rulekit.params.Measures` = \
             :class:`rulekit.params.Measures.Correlation`
             measure used during induction; default measure is correlation

diff --git a/rulekit/survival.py b/rulekit/survival.py
@@ -9,7 +9,6 @@
 
 from ._helpers import (
     PredictionResultMapper,
-    get_rule_generator,
     create_example_set
 )
 from ._operator import BaseOperator, ExpertKnowledgeOperator, Data
@@ -21,13 +20,13 @@ class SurvivalModelsParams(BaseModel):
     """Model for validating survival models hyperparameters
     """
     survival_time_attr: Optional[str]
-    minsupp_new: Optional[int] = DEFAULT_PARAMS_VALUE['minsupp_new']
+    minsupp_new: Optional[float] = DEFAULT_PARAMS_VALUE['minsupp_new']
     max_growing: Optional[float] = DEFAULT_PARAMS_VALUE['max_growing']
     enable_pruning: Optional[bool] = DEFAULT_PARAMS_VALUE['enable_pruning']
     ignore_missing: Optional[bool] = DEFAULT_PARAMS_VALUE['ignore_missing']
     max_uncovered_fraction: Optional[float] = DEFAULT_PARAMS_VALUE['max_uncovered_fraction']
     select_best_candidate: Optional[bool] = DEFAULT_PARAMS_VALUE['select_best_candidate']
-    min_rule_covered: Optional[int] = None
+    min_rule_covered: Optional[float] = None
     complementary_conditions: Optional[bool] = DEFAULT_PARAMS_VALUE['complementary_conditions']
 
     extend_using_preferred: Optional[bool] = None
@@ -63,9 +62,9 @@ def __init__(  # pylint: disable=super-init-not-called
         survival_time_attr : str
             name of column containing survival time data (use when data passed to model is padnas
             dataframe).
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         max_growing : int = 0.0
             non-negative integer representing maximum number of conditions which can be added to
             the rule in the growing phase  (use this parameter for large datasets if execution time
@@ -88,7 +87,7 @@ def __init__(  # pylint: disable=super-init-not-called
         max_rule_count : int = 0
             Maximum number of rules to be generated (for classification data sets it applies 
             to a single class); 0 indicates no limit.
-        min_rule_covered : int = None
+        min_rule_covered : float = None
             alias to `minsupp_new`. Parameter is deprecated and will be removed in the next major
             version, use `minsupp_new`
 
@@ -239,7 +238,7 @@ class ExpertSurvivalRules(ExpertKnowledgeOperator, SurvivalRules):
     def __init__(  # pylint: disable=super-init-not-called
         self,
         survival_time_attr: str = None,
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         max_growing: int = DEFAULT_PARAMS_VALUE['max_growing'],
         enable_pruning: bool = DEFAULT_PARAMS_VALUE['enable_pruning'],
         ignore_missing: bool = DEFAULT_PARAMS_VALUE['ignore_missing'],
@@ -256,14 +255,14 @@ def __init__(  # pylint: disable=super-init-not-called
         preferred_attributes_per_rule: int = DEFAULT_PARAMS_VALUE[
             'preferred_attributes_per_rule'],
         max_rule_count: int = DEFAULT_PARAMS_VALUE['max_rule_count'],
-        min_rule_covered: Optional[int] = None
+        min_rule_covered: Optional[float] = None
     ):
         """
         Parameters
         ----------
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be
-            covered by a new rule (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         survival_time_attr : str
             name of column containing survival time data (use when data passed to model is pandas
             dataframe).
@@ -309,7 +308,7 @@ def __init__(  # pylint: disable=super-init-not-called
             maximum number of preferred conditions per rule; default: unlimited,
         preferred_attributes_per_rule : int = None
             maximum number of preferred attributes per rule; default: unlimited.
-        min_rule_covered : int = None
+        min_rule_covered : float = None
             alias to `minsupp_new`. Parameter is deprecated and will be removed in the next major
             version, use `minsupp_new`
 
@@ -420,7 +419,7 @@ def __init__(  # pylint: disable=super-init-not-called
         penalty_saturation: float = DEFAULT_PARAMS_VALUE['penalty_saturation'],
 
         survival_time_attr: str = None,
-        minsupp_new: int = DEFAULT_PARAMS_VALUE['minsupp_new'],
+        minsupp_new: float = DEFAULT_PARAMS_VALUE['minsupp_new'],
         max_growing: int = DEFAULT_PARAMS_VALUE['max_growing'],
         enable_pruning: bool = DEFAULT_PARAMS_VALUE['enable_pruning'],
         ignore_missing: bool = DEFAULT_PARAMS_VALUE['ignore_missing'],
@@ -446,9 +445,9 @@ def __init__(  # pylint: disable=super-init-not-called
         survival_time_attr : str
             name of column containing survival time data (use when data passed to model is pandas
             dataframe).
-        minsupp_new : int = 5
-            positive integer representing minimum number of previously uncovered examples to be 
-            covered by a new rule  (positive examples for classification problems); default: 5
+        minsupp_new : float = 5.0
+            a minimum number (or fraction, if value < 1.0) of previously uncovered examples 
+            to be covered by a new rule (positive examples for classification problems); default: 5,
         max_growing : int = 0.0
             non-negative integer representing maximum number of conditions which can be added to 
             the rule in the growing phase (use this parameter for large datasets if execution time

diff --git a/setup.py b/setup.py
@@ -9,7 +9,7 @@
 
 setuptools.setup(
     name="rulekit",
-    version='1.7.5',
+    version='1.7.6',
     author="Cezary Maszczyk",
     author_email="cezary.maszczyk@gmail.com",
     description="Comprehensive suite for rule-based learning",