Merge branch 'master' into gaugup/ReplaceBostonHousingDataset

Author: gaugup

.github/workflows/pythonpackage.yml

@@ -15,7 +15,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        python-version: [3.6, 3.7, 3.8, 3.9]
+        python-version: [3.6, 3.7, 3.8, 3.9, "3.10"]
         os: [ubuntu-latest, macos-latest]
         exclude:
           - os: macos-latest
@@ -47,13 +47,18 @@ jobs:
       if: always()
       with:
         files: junit/test-results.xml
-    - name: Upload code coverage results
-      uses: actions/upload-artifact@v2
+    - name: Upload coverage to Codecov
+      uses: codecov/codecov-action@v2
       with:
-        name: code-coverage-results
-        path: htmlcov
-      # Use always() to always run this step to publish test results when there are test failures
-      if: ${{ always() }}
+        token: ${{ secrets.CODECOV_TOKEN }}
+        directory: .
+        env_vars: OS,PYTHON
+        fail_ci_if_error: true
+        files: ./coverage.xml
+        flags: unittests
+        name: codecov-umbrella
+        path_to_write_report: ./coverage/codecov_report.txt
+        verbose: true
     - name: Check package consistency with twine
       run: |
         python setup.py check sdist bdist_wheel

dice_ml/explainer_interfaces/dice_KD.py

@@ -89,9 +89,6 @@ def _generate_counterfactuals(self, query_instance, total_CFs, desired_range=Non
 
         query_instance[self.data_interface.outcome_name] = test_pred
         desired_class = self.misc_init(stopping_threshold, desired_class, desired_range, test_pred)
-        if desired_range is not None:
-            if desired_range[0] > desired_range[1]:
-                raise ValueError("Invalid Range!")
 
         if desired_class == "opposite" and self.model.model_type == ModelTypes.Classifier:
             if self.num_output_nodes == 2:

dice_ml/explainer_interfaces/explainer_base.py

@@ -81,6 +81,14 @@ def _validate_counterfactual_configuration(
                 raise UserConfigValidationException(
                     'The desired_range parameter should be set for regression task')
 
+        if desired_range is not None:
+            if len(desired_range) != 2:
+                raise UserConfigValidationException(
+                    "The parameter desired_range needs to have two numbers in ascending order.")
+            if desired_range[0] > desired_range[1]:
+                raise UserConfigValidationException(
+                    "The range provided in desired_range should be in ascending order.")
+
     def generate_counterfactuals(self, query_instances, total_CFs,
                                  desired_class="opposite", desired_range=None,
                                  permitted_range=None, features_to_vary="all",
@@ -96,7 +104,8 @@ def generate_counterfactuals(self, query_instances, total_CFs,
         :param desired_class: Desired counterfactual class - can take 0 or 1. Default value
                               is "opposite" to the outcome class of query_instance for binary classification.
         :param desired_range: For regression problems. Contains the outcome range to
-                              generate counterfactuals in.
+                              generate counterfactuals in. This should be a list of two numbers in
+                              ascending order.
         :param permitted_range: Dictionary with feature names as keys and permitted range in list as values.
                                 Defaults to the range inferred from training data.
                                 If None, uses the parameters initialized in data_interface.

requirements-deeplearning.txt

@@ -1,2 +1,2 @@
 tensorflow>=1.13.0-rc1 
-torch
+torch; python_version < '3.10'

setup.py

@@ -38,6 +38,7 @@
         "Programming Language :: Python :: 3.7",
         "Programming Language :: Python :: 3.8",
         "Programming Language :: Python :: 3.9",
+        "Programming Language :: Python :: 3.10",
         "License :: OSI Approved :: MIT License",
         "Operating System :: OS Independent",
     ],

tests/test_dice_interface/test_explainer_base.py

@@ -510,6 +510,18 @@ def test_generate_counterfactuals_user_config_validations_regression(
             explainer_function(query_instances=sample_custom_query_1,
                                total_CFs=10)
 
+        with pytest.raises(
+                UserConfigValidationException,
+                match=r'The parameter desired_range needs to have two numbers in ascending order.'):
+            explainer_function(query_instances=sample_custom_query_1,
+                               total_CFs=10, desired_range=[1, 3, 4])
+
+        with pytest.raises(
+                UserConfigValidationException,
+                match=r'The range provided in desired_range should be in ascending order.'):
+            explainer_function(query_instances=sample_custom_query_1,
+                               total_CFs=10, desired_range=[4, 3])
+
     def test_global_feature_importance_error_conditions_with_insufficient_query_points(
             self, method,
             sample_custom_query_1,

tests/test_notebooks.py

@@ -4,6 +4,7 @@
 
 import os
 import subprocess
+import sys
 import tempfile
 
 import nbformat
@@ -16,7 +17,11 @@
         "DiCE_with_advanced_options.ipynb",  # requires tensorflow 1.x
         "DiCE_getting_started_feasible.ipynb",  # needs changes after latest refactor
         "Benchmarking_different_CF_explanation_methods.ipynb"
-        ]
+]
+# notebooks that don't need to run on python 3.10
+torch_notebooks_not_3_10 = [
+    "DiCE_getting_started.ipynb"
+]
 
 # Adding the dice root folder to the python path so that jupyter notebooks
 if 'PYTHONPATH' not in os.environ:
@@ -77,6 +82,11 @@ def _notebook_run(filepath):
             nb,
             marks=[pytest.mark.skip, pytest.mark.advanced],
             id=nb)
+    elif sys.version_info >= (3, 10) and nb in torch_notebooks_not_3_10:
+        param = pytest.param(
+            nb,
+            marks=[pytest.mark.skip, pytest.mark.advanced],
+            id=nb)
     else:
         param = pytest.param(nb, id=nb)
     parameter_list.append(param)