Merge pull request #165 from openml/datasetdict

changed dataset list to dict

Author: mfeurer

openml/datasets/functions.py

@@ -160,9 +160,10 @@ def _list_datasets(api_call):
     assert datasets_dict['oml:data']['@xmlns:oml'] == \
            'http://openml.org/openml'
 
-    datasets = []
+    datasets = dict()
     for dataset_ in datasets_dict['oml:data']['oml:dataset']:
-        dataset = {'did': int(dataset_['oml:did']),
+        did = int(dataset_['oml:did'])
+        dataset = {'did': did,
                    'name': dataset_['oml:name'],
                    'format': dataset_['oml:format'],
                    'status': dataset_['oml:status']}
@@ -173,9 +174,7 @@ def _list_datasets(api_call):
             if abs(int(quality['#text']) - quality['#text']) < 0.0000001:
                 quality['#text'] = int(quality['#text'])
             dataset[quality['@name']] = quality['#text']
-
-        datasets.append(dataset)
-    datasets.sort(key=lambda t: t['did'])
+        datasets[did] = dataset
 
     return datasets
 

tests/datasets/test_datasets.py

@@ -71,13 +71,7 @@ def test_get_cached_dataset_arff_not_cached(self):
                                 openml.datasets.functions._get_cached_dataset_arff,
                                 3)
 
-    def test_list_datasets(self):
-        # We can only perform a smoke test here because we test on dynamic
-        # data from the internet...
-        datasets = openml.datasets.list_datasets()
-        # 1087 as the number of datasets on openml.org
-        self.assertGreaterEqual(len(datasets), 1087)
-        for dataset in datasets:
+    def _check_dataset(self, dataset):
             self.assertEqual(type(dataset), dict)
             self.assertGreaterEqual(len(dataset), 2)
             self.assertIn('did', dataset)
@@ -87,34 +81,29 @@ def test_list_datasets(self):
             self.assertIn(dataset['status'], ['in_preparation', 'active',
                                               'deactivated'])
 
+    def test_list_datasets(self):
+        # We can only perform a smoke test here because we test on dynamic
+        # data from the internet...
+        datasets = openml.datasets.list_datasets()
+        # 1087 as the number of datasets on openml.org
+        self.assertGreaterEqual(len(datasets), 1087)
+        for did in datasets:
+            self._check_dataset(datasets[did])
+
     def test_list_datasets_by_tag(self):
         datasets = openml.datasets.list_datasets(tag='uci')
         self.assertGreaterEqual(len(datasets), 5)
-        for dataset in datasets:
-            self.assertEqual(type(dataset), dict)
-            self.assertGreaterEqual(len(dataset), 2)
-            self.assertIn('did', dataset)
-            self.assertIsInstance(dataset['did'], int)
-            self.assertIn('status', dataset)
-            self.assertTrue(is_string(dataset['status']))
-            self.assertIn(dataset['status'], ['in_preparation', 'active',
-                                              'deactivated'])
+        for did in datasets:
+            self._check_dataset(datasets[did])
 
     def test_list_datasets_paginate(self):
         size = 10
         max = 100
         for i in range(0, max, size):
-            data = openml.datasets.list_datasets(offset=i, size=size)
-            self.assertGreaterEqual(size, len(data))
-            for dataset in data:
-                self.assertEqual(type(dataset), dict)
-                self.assertGreaterEqual(len(dataset), 2)
-                self.assertIn('did', dataset)
-                self.assertIsInstance(dataset['did'], int)
-                self.assertIn('status', dataset)
-                self.assertTrue(is_string(dataset['status']))
-                self.assertIn(dataset['status'], ['in_preparation',
-                                                  'active', 'deactivated'])
+            datasets = openml.datasets.list_datasets(offset=i, size=size)
+            self.assertGreaterEqual(size, len(datasets))
+            for did in datasets:
+                self._check_dataset(datasets[did])
 
     @unittest.skip('See https://github.com/openml/openml-python/issues/149')
     def test_check_datasets_active(self):