add sort to list_evaluations

Author: sahithyaravi

openml/evaluations/functions.py

@@ -19,6 +19,7 @@ def list_evaluations(
     uploader: Optional[List] = None,
     tag: Optional[str] = None,
     per_fold: Optional[bool] = None,
+    sort: Optional[str] = None,
     output_format: str = 'object'
 ) -> Union[Dict, pd.DataFrame]:
     """
@@ -48,6 +49,9 @@ def list_evaluations(
 
     per_fold : bool, optional
 
+    sort : str, optional
+       order of sorting evaluations, ascending ("asc") or descending ("desc")
+
     output_format: str, optional (default='object')
         The parameter decides the format of the output.
         - If 'object' the output is a dict of OpenMLEvaluation objects
@@ -77,6 +81,7 @@ def list_evaluations(
                                   flow=flow,
                                   uploader=uploader,
                                   tag=tag,
+                                  sort=sort,
                                   per_fold=per_fold_str)
 
 
@@ -87,6 +92,7 @@ def _list_evaluations(
     setup: Optional[List] = None,
     flow: Optional[List] = None,
     uploader: Optional[List] = None,
+    sort: Optional[str] = None,
     output_format: str = 'object',
     **kwargs
 ) -> Union[Dict, pd.DataFrame]:
@@ -114,6 +120,9 @@ def _list_evaluations(
     kwargs: dict, optional
         Legal filter operators: tag, limit, offset.
 
+    sort : str, optional
+        order of sorting evaluations, ascending ("asc") or descending ("desc")
+
     output_format: str, optional (default='dict')
         The parameter decides the format of the output.
         - If 'dict' the output is a dict of dict
@@ -141,6 +150,8 @@ def _list_evaluations(
         api_call += "/flow/%s" % ','.join([str(int(i)) for i in flow])
     if uploader is not None:
         api_call += "/uploader/%s" % ','.join([str(int(i)) for i in uploader])
+    if sort is not None:
+        api_call += "/sort/%s" % sort
 
     return __list_evaluations(api_call, output_format=output_format)
 

tests/test_evaluations/test_evaluation_functions.py

@@ -116,3 +116,28 @@ def test_evaluation_list_per_fold(self):
         for run_id in evaluations.keys():
             self.assertIsNotNone(evaluations[run_id].value)
             self.assertIsNone(evaluations[run_id].values)
+
+    def test_evaluation_list_sort(self):
+        openml.config.server = self.production_server
+        size = 10
+        task_id = 1769
+        # Get all evaluations of the task
+        unsorted_eval = openml.evaluations.list_evaluations(
+            "predictive_accuracy", offset=0, task=[task_id])
+        # Get top 10 evaluations of the same task
+        sorted_eval = openml.evaluations.list_evaluations(
+            "predictive_accuracy", size=size, offset=0, task=[task_id], sort="desc")
+
+        sorted_output = []
+        unsorted_output = []
+        for run_id in sorted_eval.keys():
+            sorted_output.append(sorted_eval[run_id].value)
+        for run_id in unsorted_eval.keys():
+            unsorted_output.append(unsorted_eval[run_id].value)
+
+        # Check if output from sort is sorted in the right order
+        self.assertTrue(sorted(sorted_output, reverse=True) == sorted_output)
+
+        # Compare manual sorting against sorted output
+        test_output = sorted(unsorted_output, reverse=True)
+        self.assertTrue(test_output[:size] == sorted_output)