BUG: Series groupby does not include nan counts for all categorical labels (#17605)

doc/source/whatsnew/v1.0.0.rst

@@ -438,6 +438,7 @@ Groupby/resample/rolling
 - Bug in :meth:`DataFrameGroupby.agg` not able to use lambda function with named aggregation (:issue:`27519`)
 - Bug in :meth:`DataFrame.groupby` losing column name information when grouping by a categorical column (:issue:`28787`)
 - Bug in :meth:`DataFrameGroupBy.rolling().quantile()` ignoring ``interpolation`` keyword argument (:issue:`28779`)
+- Bug in :meth:`SeriesGroupBy.count` missing unobserved categories when ``observed=False`` (:issue:`17605`)
 
 Reshaping
 ^^^^^^^^^

pandas/core/groupby/generic.py

@@ -721,12 +721,13 @@ def count(self) -> Series:
         minlength = ngroups or 0
         out = np.bincount(ids[mask], minlength=minlength)
 
-        return Series(
+        result = Series(
             out,
             index=self.grouper.result_index,
             name=self._selection_name,
             dtype="int64",
         )
+        return self._reindex_output(result)
 
     def _apply_to_column_groupbys(self, func):
         """ return a pass thru """

pandas/tests/groupby/test_categorical.py

@@ -1252,3 +1252,29 @@ def test_get_nonexistent_category():
                 {"var": [rows.iloc[-1]["var"]], "val": [rows.iloc[-1]["vau"]]}
             )
         )
+
+
+@pytest.mark.parametrize("aggregation", ["sum", "mean", "min", "count"])
+@pytest.mark.parametrize("observed", [True, False])
+def test_series_groupby_on_2_categoricals_unobserved(aggregation: str, observed: bool):
+    # GH 17605
+    df = pd.DataFrame(
+        {
+            "cat_1": pd.Categorical(list("AABB"), categories=list("ABCD")),
+            "cat_2": pd.Categorical(list("AB") * 2, categories=list("ABCD")),
+            "value": [0.1] * 4,
+        }
+    )
+
+    # Expect 1 observation for each combination of categories
+    if observed:
+        expected_length = 4
+    else:
+        expected_length = len(df["cat_1"].cat.categories) * len(
+            df["cat_2"].cat.categories
+        )
+
+    series_groupby = df.groupby(["cat_1", "cat_2"], observed=observed)["value"]
+    agg = getattr(series_groupby, aggregation)
+    result = agg()
+    assert len(result) == expected_length