Merge pull request #92 from lincc-frameworks/fix/issue/88

Fix ndf.add_nested() for empty frames

src/nested_pandas/nestedframe/core.py

@@ -99,7 +99,9 @@ def add_nested(
         # Add sources to objects
         packed = packer.pack(obj, name=name, dtype=dtype)
         label = packed.name
-        return self.assign(**{f"{label}": packed})
+        new_df = self.copy()
+        new_df[label] = packed
+        return new_df
 
     def _split_query(self, expr) -> dict:
         """Splits a pandas query into multiple subqueries for nested and base layers"""

tests/nested_pandas/nestedframe/test_nestedframe.py

@@ -1,5 +1,6 @@
 import numpy as np
 import pandas as pd
+import pyarrow as pa
 import pytest
 from nested_pandas import NestedFrame
 from pandas.testing import assert_frame_equal
@@ -130,6 +131,19 @@ def test_add_nested_with_series_and_mismatched_index():
     assert pd.isna(base.loc[1]["nested"])
 
 
+def test_add_nested_for_empty_df():
+    """Test that .add_nested() works for empty frame and empty input"""
+    base = NestedFrame(data={"a": [], "b": []}, index=[])
+    nested = pd.DataFrame(data={"c": []}, index=[])
+    new_base = base.add_nested(nested, "nested")
+
+    # Check original frame is unchanged
+    assert_frame_equal(base, NestedFrame(data={"a": [], "b": []}, index=[]))
+
+    assert "nested" in new_base.columns
+    assert_frame_equal(new_base.nested.nest.to_flat(), nested.astype(pd.ArrowDtype(pa.float64())))
+
+
 def test_query():
     """Test that NestedFrame.query handles nested queries correctly"""