Refactor LuxGroupBy using multiple inheritance (#309)

* add series equality and value counts test

* black formatting

* fix old value counts test instead

* add pandas tests

* remove str from column group

* save work on groupby bugs so far

* fix merge conflict again

* add new tests and add groupby bug fixes

* remove tests for staging

* update series tests

* add back getitem

* fix merge conflicts for staging

* remove print statements

* revert Makefile

* add test for name column case

* add test to ensure column is not all None

* re-add tests and commit new groupby implementation

* make parent LuxGroupby

* reformat LuxGroupBy and fix SeriesGroupby issue

* remove pandas tests

* run black

* finish seriesgroupby metadata propagation and add test

* run black

lux/core/__init__.py

@@ -14,7 +14,7 @@
 
 import pandas as pd
 from .frame import LuxDataFrame
-from .groupby import LuxDataFrameGroupBy
+from .groupby import LuxDataFrameGroupBy, LuxSeriesGroupBy
 from .series import LuxSeries
 
 global originalDF
@@ -58,8 +58,9 @@ def setOption(overridePandas=True):
         ) = (
             pd.io.spss.DataFrame
         ) = pd.io.stata.DataFrame = pd.io.api.DataFrame = pd.core.frame.DataFrame = LuxDataFrame
-        pd.Series = pd.core.series.Series = pd.core.groupby.ops.Series = LuxSeries
+        pd.Series = pd.core.series.Series = pd.core.groupby.ops.Series = pd._testing.Series = LuxSeries
         pd.core.groupby.generic.DataFrameGroupBy = LuxDataFrameGroupBy
+        pd.core.groupby.generic.SeriesGroupBy = LuxSeriesGroupBy
     else:
         pd.DataFrame = pd.io.parsers.DataFrame = pd.core.frame.DataFrame = originalDF
         pd.Series = originalSeries

lux/core/groupby.py

@@ -1,7 +1,7 @@
 import pandas as pd
 
 
-class LuxDataFrameGroupBy(pd.core.groupby.generic.DataFrameGroupBy):
+class LuxGroupBy(pd.core.groupby.groupby.GroupBy):
 
     _metadata = [
         "_intent",
@@ -26,35 +26,42 @@ class LuxDataFrameGroupBy(pd.core.groupby.generic.DataFrameGroupBy):
     ]
 
     def __init__(self, *args, **kwargs):
-        super(LuxDataFrameGroupBy, self).__init__(*args, **kwargs)
+        super(LuxGroupBy, self).__init__(*args, **kwargs)
 
     def aggregate(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self).aggregate(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self).aggregate(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         return ret_val
 
     def _agg_general(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self)._agg_general(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self)._agg_general(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         return ret_val
 
     def _cython_agg_general(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self)._cython_agg_general(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self)._cython_agg_general(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         return ret_val
 
     def get_group(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self).get_group(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self).get_group(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         ret_val.pre_aggregated = False  # Returned LuxDataFrame isn't pre_aggregated
         return ret_val
 
     def filter(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self).filter(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self).filter(*args, **kwargs)
+        for attr in self._metadata:
+            ret_val.__dict__[attr] = getattr(self, attr, None)
+        ret_val.pre_aggregated = False  # Returned LuxDataFrame isn't pre_aggregated
+        return ret_val
+
+    def apply(self, *args, **kwargs):
+        ret_val = super(LuxGroupBy, self).apply(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         ret_val.pre_aggregated = False  # Returned LuxDataFrame isn't pre_aggregated
@@ -68,15 +75,25 @@ def apply(self, *args, **kwargs):
         return ret_val
 
     def size(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self).size(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self).size(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         return ret_val
 
     def __getitem__(self, *args, **kwargs):
-        ret_val = super(LuxDataFrameGroupBy, self).__getitem__(*args, **kwargs)
+        ret_val = super(LuxGroupBy, self).__getitem__(*args, **kwargs)
         for attr in self._metadata:
             ret_val.__dict__[attr] = getattr(self, attr, None)
         return ret_val
 
     agg = aggregate
+
+
+class LuxDataFrameGroupBy(LuxGroupBy, pd.core.groupby.generic.DataFrameGroupBy):
+    def __init__(self, *args, **kwargs):
+        super(LuxDataFrameGroupBy, self).__init__(*args, **kwargs)
+
+
+class LuxSeriesGroupBy(LuxGroupBy, pd.core.groupby.generic.SeriesGroupBy):
+    def __init__(self, *args, **kwargs):
+        super(LuxSeriesGroupBy, self).__init__(*args, **kwargs)

lux/core/series.py

@@ -226,3 +226,18 @@ def exported(self) -> Union[Dict[str, VisList], VisList]:
                 When the exported vis is from the different tabs, return a dictionary with the action name as key and selected visualizations in the VisList. -> {"Enhance": VisList(v1, v2...), "Filter": VisList(v5, v7...), ..}
         """
         return self._ldf.exported
+
+    def groupby(self, *args, **kwargs):
+        history_flag = False
+        if "history" not in kwargs or ("history" in kwargs and kwargs["history"]):
+            history_flag = True
+        if "history" in kwargs:
+            del kwargs["history"]
+        groupby_obj = super(LuxSeries, self).groupby(*args, **kwargs)
+        for attr in self._metadata:
+            groupby_obj.__dict__[attr] = getattr(self, attr, None)
+        if history_flag:
+            groupby_obj._history = groupby_obj._history.copy()
+            groupby_obj._history.append_event("groupby", *args, **kwargs)
+        groupby_obj.pre_aggregated = True
+        return groupby_obj

tests/test_groupby.py

@@ -69,3 +69,11 @@ def test_get_group(global_var):
     new_df._repr_html_()
     assert new_df.history[0].name == "groupby"
     assert not new_df.pre_aggregated
+
+
+def test_series_groupby(global_var):
+    df = pytest.car_df
+    df._repr_html_()
+    new_ser = df.set_index("Brand")["Displacement"].groupby(level=0).agg("mean")
+    assert new_ser._history[0].name == "groupby"
+    assert new_ser.pre_aggregated