Create empty dataframe from Pandas DataFrame Model

pandera/api/dataframe/model.py

@@ -40,7 +40,7 @@
 from pandera.engines import PYDANTIC_V2
 from pandera.errors import SchemaInitError
 from pandera.import_utils import strategy_import_error
-from pandera.typing import AnnotationInfo
+from pandera.typing import AnnotationInfo, DataFrame
 from pandera.typing.common import DataFrameBase
 from pandera.utils import docstring_substitution
 
@@ -569,6 +569,13 @@ def to_json_schema(cls):
         """Serialize schema metadata into json-schema format."""
         raise NotImplementedError
 
+    @classmethod
+    def empty(
+        cls: Type[TDataFrameModel], *_args
+    ) -> DataFrame[TDataFrameModel]:
+        """Create an empty DataFrame instance."""
+        raise NotImplementedError
+
     if PYDANTIC_V2:
 
         @classmethod

pandera/api/pandas/model.py

@@ -1,5 +1,7 @@
 """Class-based api for pandas models."""
 
+import copy
+import sys
 from typing import Any, Dict, List, Optional, Tuple, Type, Union
 
 import pandas as pd
@@ -14,7 +16,18 @@
 from pandera.api.parsers import Parser
 from pandera.engines.pandas_engine import Engine
 from pandera.errors import SchemaInitError
-from pandera.typing import get_index_types, get_series_types, AnnotationInfo
+from pandera.typing import (
+    get_index_types,
+    get_series_types,
+    AnnotationInfo,
+    DataFrame,
+)
+
+# if python version is < 3.11, import Self from typing_extensions
+if sys.version_info < (3, 11):
+    from typing_extensions import Self
+else:
+    from typing import Self
 
 SchemaIndex = Union[Index, MultiIndex]
 
@@ -191,6 +204,14 @@
             },
         }
 
+    @classmethod
+    def empty(cls: Type[Self], *_args) -> DataFrame[Self]:
+        """Create an empty DataFrame with the schema of this model."""
+        schema = copy.deepcopy(cls.to_schema())
+        schema.coerce = True
+        empty_df = schema.coerce_dtype(pd.DataFrame(columns=[*schema.columns]))
+        return DataFrame[Self](empty_df)
+
 
 def _build_schema_index(
     indices: List[Index], **multiindex_kwargs: Any

tests/core/test_model.py

@@ -1542,3 +1542,17 @@ class Schema(pa.DataFrameModel):
     assert Schema.validate(df).equals(  # type: ignore [attr-defined]
         pd.DataFrame({"a": [1.0], "b": [1], "c": ["1"]})
     )
+
+
+def test_empty() -> None:
+    """Test to generate an empty DataFrameModel."""
+
+    class Schema(pa.DataFrameModel):
+        a: Series[pa.Float]
+        b: Series[pa.Int]
+        c: Series[pa.String]
+        d: Series[pa.DateTime]
+
+    df = Schema.empty()
+    assert df.empty
+    assert Schema.validate(df).empty  # type: ignore [attr-defined]