Merge pull request #546 from EducationalTestingService/update-examples

Update examples, and create tests
EducationalTestingService · Oct 16, 2019 · 4cfa233 · 4cfa233
2 parents 4c443d4 + 78dcf2c
commit 4cfa233
Show file tree

Hide file tree

Showing 42 changed files with 1,914 additions and 16 deletions.
diff --git a/.travis.yml b/.travis.yml
@@ -9,7 +9,7 @@ env:
   - TESTFILES="tests/test_featureset.py tests/test_utilities.py"
   - TESTFILES="tests/test_output.py"
   - TESTFILES="tests/test_regression.py"
-  - TESTFILES="tests/test_input.py tests/test_preprocessing.py tests/test_metrics.py tests/test_custom_learner.py tests/test_logutils.py"
+  - TESTFILES="tests/test_input.py tests/test_preprocessing.py tests/test_metrics.py tests/test_custom_learner.py tests/test_logutils.py tests/test_examples.py"
   - TESTFILES="tests/test_classification.py tests/test_cv.py tests/test_ablation.py"
 
 # run on the new Travis infrastructure

diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/make_titanic_example_data.py b/examples/make_titanic_example_data.py
@@ -51,18 +51,16 @@ def main():
         logger.info('Creating titanic/test directory')
         os.makedirs('titanic/test')
 
+    usecols_train = features_to_keep + ['PassengerId', 'Survived']
+    usecols_test = features_to_keep + ['PassengerId']
+
     # Read and write training FeatureSet
     train_fs = Reader.for_path('train.csv',
                                label_col='Survived',
                                id_col='PassengerId',
-                               quiet=False,
                                drop_blanks=True,
-                               pandas_kwargs={'usecols': ['Sex', 'Age',
-                                                          'Pclass', 'Fare',
-                                                          'SibSp', 'Parch',
-                                                          'Embarked',
-                                                          'PassengerId',
-                                                          'Survived']},
+                               pandas_kwargs={'usecols': usecols_train},
+                               quiet=False,
                                sparse=False).read()
 
     train_fs.filter(features=features_to_keep)
@@ -98,13 +96,9 @@ def main():
     # Read and write test FeatureSet
     test_fs = Reader.for_path('test.csv',
                               label_col='Survived',
-                              quiet=False,
                               drop_blanks=True,
-                              pandas_kwargs={'usecols': ['Sex', 'Age',
-                                                         'Pclass', 'Fare',
-                                                         'SibSp', 'Parch',
-                                                         'Embarked',
-                                                         'PassengerId']},
+                              pandas_kwargs={'usecols': usecols_test},
+                              quiet=False,
                               sparse=False).read()
 
     test_fs.filter(features=features_to_keep)

diff --git a/examples/titanic/evaluate_tuned.cfg b/examples/titanic/evaluate_tuned.cfg
@@ -19,6 +19,7 @@ objectives = ['accuracy']
 [Output]
 # again, these can be absolute paths
 metrics = ['roc_auc']
+probability = true
 log = output
 results = output
 predictions = output

diff --git a/examples/titanic/learning_curve.cfg b/examples/titanic/learning_curve.cfg
@@ -14,9 +14,9 @@ id_col = PassengerId
 
 [Tuning]
 grid_search = false
-objectives = ['accuracy']
 
 [Output]
 # again, these can be absolute paths
 log = output
 results = output
+metrics = ['accuracy']