to_pandas() with multy headers

src/datachain/lib/dc.py

@@ -10,6 +10,7 @@
     Union,
 )
 
+import pandas as pd
 import sqlalchemy
 
 from datachain.lib.feature import Feature, FeatureType
@@ -36,7 +37,6 @@
 from datachain.query.schema import Column, DatasetRow
 
 if TYPE_CHECKING:
-    import pandas as pd
     from typing_extensions import Self
 
     from datachain.catalog import Catalog
@@ -739,6 +739,15 @@ def from_pandas(  # type: ignore[override]
 
         return cls.from_features(name, session, **fr_map)
 
+    def to_pandas(self, levels: bool = True) -> "pd.DataFrame":
+        if not levels:
+            return super().to_pandas()
+
+        headers = self.signals_schema.get_headers()
+        transposed_result = list(map(list, zip(*self.results())))
+        data = {tuple(n): val for n, val in zip(headers, transposed_result)}
+        return pd.DataFrame(data)
+
     def parse_tabular(
         self,
         output: Optional[dict[str, FeatureType]] = None,

src/datachain/lib/signal_schema.py

@@ -309,6 +309,16 @@ def print_tree(self, indent: int = 4, start_at: int = 0):
                     sub_schema = SignalSchema({"* list of": args[0]})
                     sub_schema.print_tree(indent=indent, start_at=total_indent + indent)
 
+    def get_headers(self):
+        paths = [
+            path for path, _, has_subtree, _ in self.get_flat_tree() if not has_subtree
+        ]
+        max_length = max(len(path) for path in paths)
+        return [
+            path + [""] * (max_length - len(path)) if len(path) < max_length else path
+            for path in paths
+        ]
+
     @staticmethod
     def _type_to_str(type_):
         if get_origin(type_) == Union:

tests/unit/lib/test_datachain.py

@@ -784,3 +784,7 @@ def test_extend_features(catalog):
 
     res = dc._extend_features("sum", "num")
     assert res == sum(range(len(features)))
+
+
+def test_to_pandas():
+    pass