Added a docstring explaining why the ASV bench with custom date format was improved: the date_format parameter is not taken into account today.

Author: Sylvain MARIE

asv_bench/benchmarks/io/csv.py

@@ -109,12 +109,16 @@ def setup(self, obs, fq):
         rng = period_range(start="2000-01-01", periods=obs, freq=fq)
         self.data = DataFrame(rng)
 
-    def time_frame_period_formatting(self, obs, fq):
-        self.data.to_csv(self.fname, date_format="%Y-%m-%d___%H:%M:%S")
-
     def time_frame_period_no_format(self, obs, fq):
         self.data.to_csv(self.fname)
 
+    def time_frame_period_formatting(self, obs, fq):
+        """IMPORTANT: date_format is not actually taken into account here today
+        So the behaviour and perf is identical to `time_frame_period_no_format` above.
+        Not sure if this is a bug or a desired feature. Keeping the ASV for reference
+        for now"""
+        self.data.to_csv(self.fname, date_format="%Y-%m-%d___%H:%M:%S")
+
 
 class ToCSVPeriodIndex(BaseIO):
     fname = "__test__.csv"