Pandas updates and get_obs method

hydropandas/io/bro.py

@@ -148,14 +148,12 @@ def get_bro_groundwater(bro_id, tube_nr=None, only_metadata=False, **kwargs):
             empty_df = pd.DataFrame()
             return empty_df, meta
 
+        dfl = []
         for i, gld_id in enumerate(gld_ids):
-            if i == 0:
-                df, meta_new = measurements_from_gld(gld_id, **kwargs)
-                meta.update(meta_new)
-            else:
-                df_new, meta_new = measurements_from_gld(gld_id, **kwargs)
-                df = pd.concat([df, df_new], axis=1)
-                meta.update(meta_new)
+            df, meta_new = measurements_from_gld(gld_id, **kwargs)
+            meta.update(meta_new)
+            dfl.append(df)
+        df = pd.concat(dfl, axis=0)
 
         return df, meta
 
@@ -305,15 +303,15 @@ def measurements_from_gld(
 
     # to dataframe
     df = pd.DataFrame(
-        index=pd.to_datetime(times),
+        index=pd.to_datetime(times, utc=True).tz_convert("CET"),
         data={"values": values, "qualifier": qualifiers},
     )
 
     # wintertime
     if to_wintertime:
         # remove time zone information by transforming to dutch winter time
         df.index = pd.to_datetime(df.index, utc=True).tz_localize(None) + pd.Timedelta(
-            1, unit="H"
+            1, unit="h"
         )
 
     # duplicates

hydropandas/obs_collection.py

@@ -2031,6 +2031,50 @@ def from_pastastore(
         }
         return cls(obs_df, name=pstore.name, meta=meta)
 
+    def get_obs(self, name=None, **kwargs):
+        """get an observation object from a collection
+
+        Parameters
+        ----------
+        name : str or None, optional
+            name of the observation you want to select, by default None
+        **kwargs : any metadata, value pair e.g. for a collection of GroundwaterObs:
+            tube_nr = 1 or source = 'BRO'
+
+        Returns
+        -------
+        hpd.Obs
+            Observation object from the collection.
+
+        Raises
+        ------
+        ValueError
+            If multiple observations in the collection match the given attribute values.
+        ValueError
+            If no observation in the collection match the given attribute values.
+        """
+
+        # select by name
+        if name is None:
+            selected_obs = self
+        else:
+            selected_obs = self.loc[[name]]
+
+        # select by condition
+        for key, item in kwargs.items():
+            condition = selected_obs[key] == item
+            selected_obs = selected_obs.loc[condition]
+
+        # return an Obs objet
+        if len(selected_obs) == 1:
+            return selected_obs["obs"].values[0]
+        elif len(selected_obs) == 0:
+            raise ValueError("no observations for given conditions")
+        else:
+            raise ValueError(
+                f"multiple observations for given conditions {selected_obs.index}"
+            )
+
     def to_excel(self, path, meta_sheet_name="metadata"):
         """Write an ObsCollection to an excel, the first sheet in the excel contains the
         metadata, the other tabs are the timeseries of each observation.

tests/test_002_obs_objects.py

@@ -1,19 +1,11 @@
 import numpy as np
 import pandas as pd
+import pytest
 
 import hydropandas as hpd
 
-# import sys
-# sys.path.insert(1, "..")
 
-
-# TEST_DIR = os.path.dirname(os.path.abspath(__file__))
-# PROJECT_DIR = os.path.abspath(os.path.join(TEST_DIR, os.pardir))
-# sys.path.insert(0, PROJECT_DIR)
-# os.chdir(TEST_DIR)
-
-
-def test_groundwater_obs(name="groundwaterobs_001", tube_nr=2):
+def _get_groundwater_obs(name="groundwaterobs_001", tube_nr=2):
     df = pd.DataFrame(
         index=pd.date_range("2020-1-1", "2020-1-10"),
         data={"values": np.random.rand(10)},
@@ -41,7 +33,7 @@ def test_groundwater_obs(name="groundwaterobs_001", tube_nr=2):
     return o
 
 
-def test_waterlvl_obs():
+def _get_waterlvl_obs():
     df = pd.DataFrame(
         index=pd.date_range("2020-1-1", "2020-1-10"),
         data={"values": np.random.rand(10)},
@@ -60,6 +52,16 @@ def test_waterlvl_obs():
     return o
 
 
+def _obscollection_from_list():
+    o_list = []
+    for i in range(10):
+        o_list.append(_get_groundwater_obs(name=f"groundwaterobs_00{i}", tube_nr=i))
+
+    oc = hpd.ObsCollection.from_list(o_list)
+
+    return oc
+
+
 def test_groundwater_quality_obs():
     df = pd.DataFrame(
         index=pd.date_range("2020-1-1", "2020-1-10"), data={"pH": np.random.rand(10)}
@@ -75,25 +77,15 @@ def test_groundwater_quality_obs():
     )
 
 
-def test_obscollection_from_list():
-    o_list = []
-    for i in range(10):
-        o_list.append(test_groundwater_obs(name=f"groundwaterobs_00{i}", tube_nr=i))
-
-    oc = hpd.ObsCollection.from_list(o_list)
-
-    return oc
-
-
 def test_add_meta_to_df():
-    oc = test_obscollection_from_list()
+    oc = _obscollection_from_list()
     oc.add_meta_to_df(key="all")
 
     assert "info" in oc.columns, "unexpected result for add_meta_to_df"
 
 
 def test_copy_obs():
-    o = test_groundwater_obs(name="groundwaterobs_001", tube_nr=2)
+    o = _get_groundwater_obs(name="groundwaterobs_001", tube_nr=2)
     o2 = o.copy()
 
     o.meta["hello"] = "world"
@@ -133,10 +125,10 @@ def test_convert_waterlvl_groundwater_obs():
 
 def test_merge_observations_same_timeseries():
     # base
-    o = test_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
+    o = _get_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
 
     # observation with different metadata, same time series
-    o2 = test_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
+    o2 = _get_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
     o2.iloc[:, 0] = o.iloc[:, 0]
 
     omerged = o.merge_observation(o2, merge_metadata=False)
@@ -146,7 +138,7 @@ def test_merge_observations_same_timeseries():
 
 def test_merge_observations_different_timeseries():
     # base
-    o = test_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
+    o = _get_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
 
     # observation with different time series
     o2 = o.copy()
@@ -163,7 +155,7 @@ def test_merge_observations_different_timeseries():
 
 def test_merge_overlapping():
     # base
-    o = test_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
+    o = _get_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
 
     # observation with partially overlapping time series and extra columns
     o2 = o.copy()
@@ -182,29 +174,47 @@ def test_merge_overlapping():
 
 def test_merge_errors():
     # base
-    o = test_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
+    o = _get_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
 
     # observation with partially overlapping time series and extra columns
-    o2 = test_waterlvl_obs()
+    o2 = _get_waterlvl_obs()
 
-    try:
+    with pytest.raises(TypeError):
         o.merge_observation(o2)
-    except TypeError:
-        return
-
-    raise RuntimeError("function should raise an error")
 
 
 def test_add_observation_to_oc():
-    oc = test_obscollection_from_list()
+    oc = _obscollection_from_list()
 
-    o = test_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
+    o = _get_groundwater_obs(name="groundwaterobs_010", tube_nr=10)
 
     oc.add_observation(o)
 
 
 def test_interpolate_obscollection():
-    oc = test_obscollection_from_list()
+    oc = _obscollection_from_list()
 
     xy = [[500, 11000], [9000, 18000]]
     oc.interpolate(xy)
+
+
+def test_get_obs():
+    oc = _obscollection_from_list()
+
+    # by name
+    o = oc.get_obs(name="groundwaterobs_001")
+    assert isinstance(o, hpd.GroundwaterObs)
+    assert o.name == "groundwaterobs_001"
+
+    # by attributes
+    o = oc.get_obs(monitoring_well="groundwaterobs", tube_nr=2)
+    assert isinstance(o, hpd.GroundwaterObs)
+    assert o.tube_nr == 2
+
+    # multiple observations
+    with pytest.raises(ValueError):
+        oc.get_obs(monitoring_well="groundwaterobs")
+
+    # no observations
+    with pytest.raises(ValueError):
+        oc.get_obs(monitoring_well="I do not exist")