pyexplainer version 0.1.9 with AutoSpearman

awsm-research · Apr 14, 2021 · f728c5d · f728c5d
1 parent 6ad9406
commit f728c5d
Show file tree

Hide file tree

Showing 12 changed files with 83 additions and 13 deletions.
diff --git a/.idea/misc.xml b/.idea/misc.xml
diff --git a/.idea/pyExplainer.iml b/.idea/pyExplainer.iml
diff --git a/dist/pyexplainer-0.1.9-py3-none-any.whl b/dist/pyexplainer-0.1.9-py3-none-any.whl
diff --git a/dist/pyexplainer-0.1.9.tar.gz b/dist/pyexplainer-0.1.9.tar.gz
diff --git a/poetry.lock b/poetry.lock
diff --git a/pyexplainer/__init__.py b/pyexplainer/__init__.py
@@ -1 +1 @@
-__version__ = '0.1.8'
+__version__ = '0.1.9'
diff --git a/pyexplainer/__pycache__/pyexplainer_pyexplainer.cpython-38.pyc b/pyexplainer/__pycache__/pyexplainer_pyexplainer.cpython-38.pyc
diff --git a/pyexplainer/pyexplainer_pyexplainer.py b/pyexplainer/pyexplainer_pyexplainer.py
@@ -228,23 +228,20 @@ def __init__(self,
         self.X_explain = None
         self.y_explain = None
 
-    def auto_spearman(self, correlation_threshold=0.7, correlation_method='spearman', VIF_threshold=5):
+    def auto_spearman(self, apply_to_X_train=True, correlation_threshold=0.7, correlation_method='spearman', VIF_threshold=5):
         """An automated feature selection approach that address collinearity and multicollinearity.
         For more information, please kindly refer to the `paper <https://ieeexplore.ieee.org/document/8530020>`_.
 
         Parameters
         ----------
+        apply_to_X_train : :obj:`bool`
+            Whether to apply the selected columns to the X_train data inside PyExplainer Obj., default is True
         correlation_threshold : :obj:`float`
             Threshold value of correalation.
         correlation_method : :obj:`str`
             Method for solving the correlation between the features.
         VIF_threshold : :obj:`int`
             Threshold value of VIF score.
-
-        Returns
-        -------
-        :obj:`pandas.core.frame.DataFrame`
-            A Dataframe that contains selected features
         """
         X_AS_train = self.X_train.copy()
         AS_metrics = X_AS_train.columns
@@ -336,7 +333,9 @@ def auto_spearman(self, correlation_threshold=0.7, correlation_method='spearman'
             X_AS_train = X_AS_train.loc[:, selected_features]
 
         print('Finally, according to Part 2 of AutoSpearman,', AS_metrics, 'are selected.')
-        return AS_metrics
+        if apply_to_X_train:
+            self.set_X_train(X_AS_train)
+            print('X_train data inside PyExplainer was updated based on the selected features above')
 
     def explain(self,
                 X_explain,
@@ -1291,6 +1290,13 @@ def set_top_k_rules(self, top_k_rules):
             self.top_k_rules = top_k_rules
 
     def set_X_train(self, X_train):
+        """Setter of X_train
+
+        Parameters
+        ----------
+        X_train : :obj:`pandas.core.frame.DataFrame`
+            X_train data
+        """
         if isinstance(X_train, pd.core.frame.DataFrame):
             self.X_train = X_train
         else:

diff --git a/pyproject.toml b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pyexplainer"
-version = "0.1.8"
+version = "0.1.9"
 description = "Explainable AI tool for Software Quality Assurance (SQA)"
 authors = ["Michael <michaelfu1998@gmail.com>"]
 license = "MIT"
@@ -13,6 +13,7 @@ scipy = "^1.6.1"
 ipywidgets = "^7.6.3"
 ipython = "^7.21.0"
 pandas = "^1.2.3"
+statsmodels = "^0.12.2"
 
 [tool.poetry.dev-dependencies]
 Sphinx = "^3.5.1"

diff --git a/setup.py b/setup.py
diff --git a/tests/__pycache__/test_pyexplainer.cpython-38-pytest-6.2.2.pyc b/tests/__pycache__/test_pyexplainer.cpython-38-pytest-6.2.2.pyc
diff --git a/tests/test_pyexplainer.py b/tests/test_pyexplainer.py
@@ -120,7 +120,7 @@ def in_virtualenv():
 
 
 def test_version():
-    assert __version__ == '0.1.8'
+    assert __version__ == '0.1.9'
 
 
 @pytest.mark.parametrize('data, result',
@@ -205,6 +205,11 @@ def test_pyexplainer_init_positive(X_train, y_train, indep, dep, blackbox_model,
                  'dep', 'top_k_positive_rules', 'top_k_negative_rules', 'local_rulefit_model']
 
 
+def test_auto_spearman():
+    py_explainer.auto_spearman()
+    py_explainer.auto_spearman(apply_to_X_train=False)
+
+
 @pytest.mark.parametrize('exp_X_explain, exp_y_explain, top_k, max_rules, max_iter, cv, search_function, debug, '
                          'X_train, result',
                          [