Flow Processor. Improve datetime_out computation.

cfatools/processor/flow.py

@@ -39,31 +39,38 @@ def get_flow_timeseries(dataset: DatasetReader):
     pass
 
 
-def add_datetime_out(event_df: pd.DataFrame, mlmin_df: pd.DataFrame, tube_volume_ml: float) -> pd.DataFrame:
-    """
+def get_datetime_out(datetime_in: pd.Series, mlmin_df: pd.DataFrame, tube_volume_ml: float) -> pd.Series:
+    """Get the date/time the fluid reaches the end of the tube.
+
+    Considering a tube whose internal volume is ``tube_volume_ml``, in which flows a fluid as variable flow rates
+    listed and timestamped in ``mlmin_df``, if some "transitions" (e.g. a breaks between 2 ice cores) enters the tube at
+    ``datetime_in``, then this function will return the moments these transitions reach the output of the tube.
 
     Parameters
     ----------
-    event_df: pd.DataFrame
+    datetime_in: pd.Series
+        Datetime-indexed series of moments the fluids of interest enters the tube.
     mlmin_df: pd.DataFrame
+        Datetime-indexed dataframe containing a Series of fluid flow rate, in ml/min (data column must be named
+        ``mlmin``). In most case, this comes from pump's flow rate settings time series.
     tube_volume_ml: float
 
     Returns
     -------
-    pd.DataFrame
-
-    """
+    pd.Series
+        Date/time when the fluid of interest exists the tube.
+    """    
     # Add events' datetime to pump's mlmin change datetime
-    mlmin_df = pd.merge(mlmin_df, event_df, left_index=True, right_index=True, how="outer")
+    mlmin_df = pd.merge(mlmin_df, datetime_in, left_index=True, right_index=True, how="outer")
     mlmin_df = mlmin_df[["mlmin"]].ffill()
     mlmin_df["ml_cumul"] = pump.add_cumulated_volume(mlmin_df["mlmin"])
 
-    event_df["datetime_out"] = event_df["datetime_in"].apply(get_datetime_out, args=(mlmin_df, tube_volume_ml))
+    datetime_out = datetime_in.apply(__get_single_datetime_out__, args=(mlmin_df, tube_volume_ml))
 
-    return event_df
+    return datetime_out
 
 
-def get_datetime_out(datetime_in: pd.Timestamp, mlmin_df: pd.DataFrame, tube_volume_ml: float) -> pd.Timestamp:
+def __get_single_datetime_out__(datetime_in: pd.Timestamp, mlmin_df: pd.DataFrame, tube_volume_ml: float) -> pd.Timestamp:
     """Get the date/time the fluid reaches the end of the tube.
 
     Considering a tube whose internal volume is ``tube_volume_ml``, in which flows a fluid as variable flow rates

cfatools/tests/test_processor/test_flow.py

@@ -28,7 +28,7 @@ class TestFlow(TestCase):
         except:
             self.fail("Exception raised!")
 
-    def test_get_datetime_out(self):
+    def test_get_datetime_out_artificialdataset(self):
         tube_volume_ml = 30
 
         # Build variable flow rate test set
@@ -49,23 +49,15 @@ class TestFlow(TestCase):
 
         icecore_df.set_index("datetime_in", inplace=True)
         icecore_df["datetime_in"] = icecore_df.index
-        # icecore_df["datetime_out"] = icecore_df["datetime_in"].apply(flow.get_datetime_out,
-        #                                                              args=(mlmin_df, tube_volume_ml))
-        flow.add_datetime_out(icecore_df, mlmin_df, tube_volume_ml)
-        # try:
-        #
-        # except:
-        #     self.fail("Exception raised!")
+
+        icecore_df["datetime_out"] = flow.get_datetime_out(icecore_df["datetime_in"], mlmin_df, tube_volume_ml)
 
         self.assertEqual(icecore_df.iloc[0]["datetime_out"], pd.to_datetime("2021-01-01 00:25:00"))
         self.assertEqual(icecore_df.iloc[1]["datetime_out"], pd.to_datetime("2021-01-01 00:29:00"))
         self.assertEqual(icecore_df.iloc[2]["datetime_out"], pd.to_datetime("2021-01-01 00:30:00"))
         self.assertEqual(icecore_df.iloc[3]["datetime_out"], pd.to_datetime("2021-01-01 00:32:00"))
 
-        ########################
-        # Real dataset
-
-    def test_add_datetime_out(self):
+    def test_get_datetime_out_realdateset(self):
         tube_volume_ml = 30
 
         dataset_name = "20210122_mock_ice_bronkgas_debub"
@@ -75,20 +67,13 @@ class TestFlow(TestCase):
 
         encoder_df["datetime_in"] = encoder_df.index
 
-        t0 = datetime.datetime.now()
-
-
-        t1 = datetime.datetime.now()
         profile = cProfile.Profile()
         profile.enable()
 
-        encoder_df = flow.add_datetime_out(encoder_df, colpump_df, tube_volume_ml)
+        encoder_df["datetime_out"] = flow.get_datetime_out(encoder_df["datetime_in"], colpump_df, tube_volume_ml)
 
         profile.disable()
         ps = pstats.Stats(profile)
         ps.print_stats()
 
-        t2 = datetime.datetime.now()
-        print(t1 - t0)
 
-        print(t2 - t1)