BUG: pd.DataFrame.from_records() raises a KeyError if passed a string index and an empty iterable

doc/source/whatsnew/v1.5.0.rst

@@ -776,7 +776,7 @@ Conversion
 Strings
 ^^^^^^^
 - Bug in :meth:`str.startswith` and :meth:`str.endswith` when using other series as parameter _pat_. Now raises ``TypeError`` (:issue:`3485`)
--
+- Bug in :meth:`DataFrame.from_records` raises a ``KeyError`` if passed a string field label and an empty iterable (:isue:`#47285`)
 
 Interval
 ^^^^^^^^

pandas/core/frame.py

@@ -2291,6 +2291,14 @@ def maybe_reorder(
             arr_columns = ensure_index(arr_columns)
             if columns is None:
                 columns = arr_columns
+
+                if len(arrays) == 0:
+                    if index is not None:
+                        if isinstance(index, str) or not hasattr(index, "__iter__"):
+                            if index not in columns:
+                                result_index = [index]
+                                index = None
+
             else:
                 arrays, arr_columns, result_index = maybe_reorder(
                     arrays, arr_columns, columns, index

pandas/tests/io/formats/test_info.py

@@ -491,3 +491,20 @@ def test_info_int_columns():
         """
     )
     assert result == expected
+
+
+def test_info_from_rec():
+    # GH47285
+    df = DataFrame.from_records([], index="foo")
+    buf = StringIO()
+    df.info(buf=buf)
+
+    itr = [
+        {"col_1": 3, "col_2": "a"},
+        {"col_1": 2, "col_2": "b"},
+        {"col_1": 1, "col_2": "c"},
+        {"col_1": 0, "col_2": "d"},
+    ]
+    with pytest.raises(KeyError) as err:
+        df = DataFrame.from_records(itr, index="col_14")
+    assert "col_14" in str(err.value)