maxframe 2.2.0__cp312-cp312-macosx_10_9_universal2.whl → 2.3.0rc1__cp312-cp312-macosx_10_9_universal2.whl

maxframe/dataframe/core.py

@@ -715,55 +715,6 @@ class IndexData(HasShapeTileableData, _ToPandasMixin):
         return from_index(self, dtype=dtype, extract_multi_index=extract_multi_index)
 
     def to_frame(self, index: bool = True, name=None):
-        """
-        Create a DataFrame with a column containing the Index.
-
-        Parameters
-        ----------
-        index : bool, default True
-            Set the index of the returned DataFrame as the original Index.
-
-        name : object, default None
-            The passed name should substitute for the index name (if it has
-            one).
-
-        Returns
-        -------
-        DataFrame
-            DataFrame containing the original Index data.
-
-        See Also
-        --------
-        Index.to_series : Convert an Index to a Series.
-        Series.to_frame : Convert Series to DataFrame.
-
-        Examples
-        --------
-        >>> import maxframe.dataframe as md
-        >>> idx = md.Index(['Ant', 'Bear', 'Cow'], name='animal')
-        >>> idx.to_frame().execute()
-               animal
-        animal
-        Ant       Ant
-        Bear     Bear
-        Cow       Cow
-
-        By default, the original Index is reused. To enforce a new Index:
-
-        >>> idx.to_frame(index=False).execute()
-          animal
-        0    Ant
-        1   Bear
-        2    Cow
-
-        To override the name of the resulting column, specify `name`:
-
-        >>> idx.to_frame(index=False, name='zoo').execute()
-            zoo
-        0   Ant
-        1  Bear
-        2   Cow
-        """
         from . import dataframe_from_tensor
 
         if isinstance(self.index_value.value, IndexValue.MultiIndex):
@@ -789,34 +740,20 @@ class IndexData(HasShapeTileableData, _ToPandasMixin):
             columns = [name or self.name or 0]
         index_ = self if index else None
         return dataframe_from_tensor(
-            self._to_maxframe_tensor(self, extract_multi_index=True),
+            self._to_maxframe_tensor(extract_multi_index=True),
             index=index_,
             columns=columns,
         )
 
     def to_series(self, index=None, name=None):
-        """
-        Create a Series with both index and values equal to the index keys.
-
-        Useful with map for returning an indexer based on an index.
-
-        Parameters
-        ----------
-        index : Index, optional
-            Index of resulting Series. If None, defaults to original index.
-        name : str, optional
-            Dame of resulting Series. If None, defaults to name of original
-            index.
-
-        Returns
-        -------
-        Series
-            The dtype will be based on the type of the Index values.
-        """
         from . import series_from_index
 
         return series_from_index(self, index=index, name=name)
 
+    @property
+    def hasnans(self):
+        return self.isna().any()
+
 
 class Index(HasShapeTileable, _ToPandasMixin):
     __slots__ = "_df_or_series", "_parent_key", "_axis"
@@ -887,6 +824,99 @@ class Index(HasShapeTileable, _ToPandasMixin):
     def values(self):
         return self.to_tensor()
 
+    def to_frame(self, index: bool = True, name=None):
+        """
+        Create a DataFrame with a column containing the Index.
+
+        Parameters
+        ----------
+        index : bool, default True
+            Set the index of the returned DataFrame as the original Index.
+
+        name : object, default None
+            The passed name should substitute for the index name (if it has
+            one).
+
+        Returns
+        -------
+        DataFrame
+            DataFrame containing the original Index data.
+
+        See Also
+        --------
+        Index.to_series : Convert an Index to a Series.
+        Series.to_frame : Convert Series to DataFrame.
+
+        Examples
+        --------
+        >>> import maxframe.dataframe as md
+        >>> idx = md.Index(['Ant', 'Bear', 'Cow'], name='animal')
+        >>> idx.to_frame().execute()
+               animal
+        animal
+        Ant       Ant
+        Bear     Bear
+        Cow       Cow
+
+        By default, the original Index is reused. To enforce a new Index:
+
+        >>> idx.to_frame(index=False).execute()
+          animal
+        0    Ant
+        1   Bear
+        2    Cow
+
+        To override the name of the resulting column, specify `name`:
+
+        >>> idx.to_frame(index=False, name='zoo').execute()
+            zoo
+        0   Ant
+        1  Bear
+        2   Cow
+        """
+        return self._data.to_frame(index=index, name=name)
+
+    def to_series(self, index=None, name=None):
+        """
+        Create a Series with both index and values equal to the index keys.
+
+        Useful with map for returning an indexer based on an index.
+
+        Parameters
+        ----------
+        index : Index, optional
+            Index of resulting Series. If None, defaults to original index.
+        name : str, optional
+            Dame of resulting Series. If None, defaults to name of original
+            index.
+
+        Returns
+        -------
+        Series
+            The dtype will be based on the type of the Index values.
+        """
+        return self._data.to_series(index=index, name=name)
+
+    @property
+    def hasnans(self):
+        """
+        Return True if there are any NaNs.
+
+        Returns
+        -------
+        bool
+
+        Examples
+        --------
+        >>> import maxframe.dataframe as md
+        >>> idx = md.Index([1, 2, 3, None])
+        >>> idx.execute()
+        Index([1.0, 2.0, 3.0, nan], dtype='float64')
+        >>> idx.hasnans.execute()
+        True
+        """
+        return self._data.hasnans
+
 
 class RangeIndex(Index):
     __slots__ = ()
@@ -1085,12 +1115,6 @@ class SeriesData(_BatchedFetcher, BaseSeriesData):
 
     items = iteritems
 
-    def to_dict(self, into=dict, batch_size=10000, session=None):
-        fetch_kwargs = dict(batch_size=batch_size)
-        return self.to_pandas(session=session, fetch_kwargs=fetch_kwargs).to_dict(
-            into=into
-        )
-
     def to_frame(self, name=None):
         from . import dataframe_from_tensor
 
@@ -1285,38 +1309,6 @@ class Series(HasShapeTileable, _ToPandasMixin):
 
     items = iteritems
 
-    def to_dict(self, into=dict, batch_size=10000, session=None):
-        """
-        Convert Series to {label -> value} dict or dict-like object.
-
-        Parameters
-        ----------
-        into : class, default dict
-            The collections.abc.Mapping subclass to use as the return
-            object. Can be the actual class or an empty
-            instance of the mapping type you want.  If you want a
-            collections.defaultdict, you must pass it initialized.
-
-        Returns
-        -------
-        collections.abc.Mapping
-            Key-value representation of Series.
-
-        Examples
-        --------
-        >>> import maxframe.dataframe as md
-        >>> s = md.Series([1, 2, 3, 4])
-        >>> s.to_dict()
-        {0: 1, 1: 2, 2: 3, 3: 4}
-        >>> from collections import OrderedDict, defaultdict
-        >>> s.to_dict(OrderedDict)
-        OrderedDict([(0, 1), (1, 2), (2, 3), (3, 4)])
-        >>> dd = defaultdict(list)
-        >>> s.to_dict(dd)
-        defaultdict(<class 'list'>, {0: 1, 1: 2, 2: 3, 3: 4})
-        """
-        return self._data.to_dict(into=into, batch_size=batch_size, session=session)
-
     def to_frame(self, name=None):
         """
         Convert Series to DataFrame.

maxframe/dataframe/datasource/core.py

@@ -18,6 +18,7 @@ from typing import List, MutableMapping, Optional, Union
 from ...serialization.serializables import Int64Field, StringField
 from ...utils import estimate_pandas_size
 from ..operators import DataFrameOperator, DataFrameOperatorMixin
+from ..utils import validate_dtype_backend
 
 
 class HeadOptimizedDataSource(DataFrameOperator, DataFrameOperatorMixin):
@@ -86,3 +87,8 @@ class PandasDataSourceOperator(DataFrameOperator):
         cls, ctx: MutableMapping[str, Union[int, float]], op: "PandasDataSourceOperator"
     ):
         ctx[op.outputs[0].key] = estimate_pandas_size(op.get_data())
+
+
+class DtypeBackendCompatibleMixin:
+    def __on_deserialize__(self):
+        self.dtype_backend = validate_dtype_backend(self.dtype_backend)

maxframe/dataframe/datasource/direct.py

@@ -0,0 +1,57 @@
+# Copyright 1999-2025 Alibaba Group Holding Ltd.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import pandas as pd
+
+
+def read_clipboard(sep=None, **kwargs):
+    """
+    Read text from clipboard and pass to :func:`~pandas.read_csv`.
+
+    Parses clipboard contents similar to how CSV files are parsed
+    using :func:`~pandas.read_csv`.
+
+    Parameters
+    ----------
+    sep : str, default '\\s+'
+        A string or regex delimiter. The default of ``'\\s+'`` denotes
+        one or more whitespace characters.
+
+    **kwargs
+        See :func:`~pandas.read_csv` for the full argument list.
+
+    Returns
+    -------
+    DataFrame
+        A parsed :class:`DataFrame` object.
+
+    See Also
+    --------
+    DataFrame.to_clipboard : Copy object to the system clipboard.
+    read_csv : Read a comma-separated values (csv) file into DataFrame.
+    read_fwf : Read a table of fixed-width formatted lines into DataFrame.
+
+    Examples
+    --------
+    >>> import maxframe.dataframe as md
+    >>> df = md.DataFrame([[1, 2, 3], [4, 5, 6]], columns=['A', 'B', 'C'])
+    >>> df.to_clipboard()  # doctest: +SKIP
+    >>> md.read_clipboard()  # doctest: +SKIP.execute()
+         A  B  C
+    0    1  2  3
+    1    4  5  6
+    """
+    from ..initializer import DataFrame
+
+    return DataFrame(pd.read_clipboard(sep=sep, **kwargs))

maxframe/dataframe/datasource/read_csv.py

@@ -38,8 +38,12 @@ from ...serialization.serializables import (
     StringField,
 )
 from ...utils import lazy_import, parse_readable_size
-from ..utils import parse_index, to_arrow_dtypes
-from .core import ColumnPruneSupportedDataSourceMixin, IncrementalIndexDatasource
+from ..utils import parse_index, to_arrow_dtypes, validate_dtype_backend
+from .core import (
+    ColumnPruneSupportedDataSourceMixin,
+    DtypeBackendCompatibleMixin,
+    IncrementalIndexDatasource,
+)
 
 cudf = lazy_import("cudf")
 
@@ -88,6 +92,7 @@ def _find_chunk_start_end(f, offset, size):
 class DataFrameReadCSV(
     IncrementalIndexDatasource,
     ColumnPruneSupportedDataSourceMixin,
+    DtypeBackendCompatibleMixin,
 ):
     _op_type_ = opcodes.READ_CSV
 
@@ -101,7 +106,7 @@ class DataFrameReadCSV(
     offset = Int64Field("offset")
     size = Int64Field("size")
     incremental_index = BoolField("incremental_index")
-    use_arrow_dtype = BoolField("use_arrow_dtype")
+    dtype_backend = StringField("dtype_backend", default=None)
     keep_usecols_order = BoolField("keep_usecols_order", default=None)
     storage_options = DictField("storage_options")
     merge_small_files = BoolField("merge_small_files")
@@ -151,7 +156,7 @@ def read_csv(
     head_bytes="100k",
     head_lines=None,
     incremental_index: bool = True,
-    use_arrow_dtype: bool = None,
+    dtype_backend: str = None,
     storage_options: dict = None,
     memory_scale: int = None,
     merge_small_files: bool = True,
@@ -419,8 +424,8 @@ def read_csv(
     incremental_index: bool, default True
         If index_col not specified, ensure range index incremental,
         gain a slightly better performance if setting False.
-    use_arrow_dtype: bool, default None
-        If True, use arrow dtype to store columns.
+    dtype_backend: {'numpy', 'pyarrow'}, default 'numpy'
+        Back-end data type applied to the resultant DataFrame (still experimental).
     storage_options: dict, optional
         Options for storage connection.
     merge_small_files: bool, default True
@@ -509,7 +514,7 @@ def read_csv(
         compression=compression,
         gpu=gpu,
         incremental_index=incremental_index,
-        use_arrow_dtype=use_arrow_dtype,
+        dtype_backend=dtype_backend,
         storage_options=storage_options,
         memory_scale=memory_scale,
         merge_small_files=merge_small_files,
@@ -518,10 +523,13 @@ def read_csv(
     )
     chunk_bytes = chunk_bytes or options.chunk_store_limit
     dtypes = mini_df.dtypes
-    if use_arrow_dtype is None:
-        use_arrow_dtype = options.dataframe.use_arrow_dtype
-    if not gpu and use_arrow_dtype:
-        dtypes = to_arrow_dtypes(dtypes, test_df=mini_df)
+
+    dtype_backend = validate_dtype_backend(
+        dtype_backend or options.dataframe.dtype_backend
+    )
+
+    if not gpu and dtype_backend == "pyarrow":
+        dtypes = to_arrow_dtypes(dtypes)
     ret = op(
         index_value=index_value,
         columns_value=columns_value,

maxframe/dataframe/datasource/read_odps_query.py

@@ -29,7 +29,7 @@ from odps.types import Column, OdpsSchema, validate_data_type
 from odps.utils import split_sql_by_semicolon
 
 from ... import opcodes
-from ...config import options
+from ...config import option_context, options
 from ...core import OutputType
 from ...core.graph import DAG
 from ...io.odpsio import odps_schema_to_pandas_dtypes
@@ -44,8 +44,12 @@ from ...serialization.serializables import (
     StringField,
 )
 from ...utils import is_empty
-from ..utils import parse_index
-from .core import ColumnPruneSupportedDataSourceMixin, IncrementalIndexDatasource
+from ..utils import parse_index, validate_dtype_backend
+from .core import (
+    ColumnPruneSupportedDataSourceMixin,
+    DtypeBackendCompatibleMixin,
+    IncrementalIndexDatasource,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -266,6 +270,7 @@ def _build_explain_sql(
 class DataFrameReadODPSQuery(
     IncrementalIndexDatasource,
     ColumnPruneSupportedDataSourceMixin,
+    DtypeBackendCompatibleMixin,
 ):
     _op_type_ = opcodes.READ_ODPS_QUERY
 
@@ -273,12 +278,16 @@ class DataFrameReadODPSQuery(
     dtypes = SeriesField("dtypes", default=None)
     columns = AnyField("columns", default=None)
     nrows = Int64Field("nrows", default=None)
-    use_arrow_dtype = BoolField("use_arrow_dtype", default=None)
+    dtype_backend = StringField("dtype_backend", default=None)
     string_as_binary = BoolField("string_as_binary", default=None)
     index_columns = ListField("index_columns", FieldTypes.string, default=None)
     index_dtypes = SeriesField("index_dtypes", default=None)
     column_renames = DictField("column_renames", default=None)
 
+    def __init__(self, dtype_backend=None, **kw):
+        dtype_backend = validate_dtype_backend(dtype_backend)
+        super().__init__(dtype_backend=dtype_backend, **kw)
+
     def get_columns(self):
         return self.columns or list(self.dtypes.index)
 
@@ -404,6 +413,7 @@ def read_odps_query(
     sql_hints: Dict[str, str] = None,
     anonymous_col_prefix: str = _DEFAULT_ANONYMOUS_COL_PREFIX,
     skip_schema: bool = False,
+    dtype_backend: str = None,
     **kw,
 ):
     """
@@ -428,6 +438,8 @@ def read_odps_query(
         Skip resolving output schema before execution. Once this is configured,
         the output DataFrame cannot be inputs of other DataFrame operators
         before execution.
+    dtype_backend: {'numpy', 'pyarrow'}, default 'numpy'
+        Back-end data type applied to the resultant DataFrame (still experimental).
 
     Returns
     -------
@@ -459,6 +471,14 @@ def read_odps_query(
     if odps_entry is None:
         raise ValueError("Missing odps_entry parameter")
 
+    if "use_arrow_dtype" in kw:
+        dtype_backend = dtype_backend or validate_dtype_backend(
+            kw.pop("use_arrow_dtype")
+        )
+    dtype_backend = validate_dtype_backend(
+        dtype_backend or options.dataframe.dtype_backend
+    )
+
     col_renames = {}
     if not skip_schema:
         odps_schema = _resolve_query_schema(
@@ -479,7 +499,9 @@ def read_odps_query(
             else:
                 new_columns.append(col)
 
-        dtypes = odps_schema_to_pandas_dtypes(OdpsSchema(new_columns))
+        with option_context():
+            options.dataframe.dtype_backend = dtype_backend
+            dtypes = odps_schema_to_pandas_dtypes(OdpsSchema(new_columns))
     else:
         dtypes = None
 
@@ -500,10 +522,11 @@ def read_odps_query(
 
     chunk_bytes = kw.pop("chunk_bytes", None)
     chunk_size = kw.pop("chunk_size", None)
+
     op = DataFrameReadODPSQuery(
         query=query,
         dtypes=dtypes,
-        use_arrow_dtype=kw.pop("use_arrow_dtype", True),
+        dtype_backend=dtype_backend,
         string_as_binary=string_as_binary,
         index_columns=index_col,
         index_dtypes=index_dtypes,

maxframe/dataframe/datasource/read_odps_table.py

@@ -22,7 +22,7 @@ from odps.models import Table
 from odps.utils import to_timestamp
 
 from ... import opcodes
-from ...config import options
+from ...config import option_context, options
 from ...core import OutputType
 from ...io.odpsio import odps_schema_to_pandas_dtypes
 from ...serialization.serializables import (
@@ -36,8 +36,12 @@ from ...serialization.serializables import (
 )
 from ...utils import estimate_table_size, is_empty
 from ..core import DataFrame  # noqa: F401
-from ..utils import parse_index
-from .core import ColumnPruneSupportedDataSourceMixin, IncrementalIndexDatasource
+from ..utils import parse_index, validate_dtype_backend
+from .core import (
+    ColumnPruneSupportedDataSourceMixin,
+    DtypeBackendCompatibleMixin,
+    IncrementalIndexDatasource,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -45,6 +49,7 @@ logger = logging.getLogger(__name__)
 class DataFrameReadODPSTable(
     IncrementalIndexDatasource,
     ColumnPruneSupportedDataSourceMixin,
+    DtypeBackendCompatibleMixin,
 ):
     __slots__ = ("_odps_entry",)
     _op_type_ = opcodes.READ_ODPS_TABLE
@@ -54,18 +59,22 @@ class DataFrameReadODPSTable(
     dtypes = SeriesField("dtypes", default=None)
     columns = AnyField("columns", default=None)
     nrows = Int64Field("nrows", default=None)
-    use_arrow_dtype = BoolField("use_arrow_dtype", default=None)
+    dtype_backend = StringField("dtype_backend", default=None)
     string_as_binary = BoolField("string_as_binary", default=None)
     append_partitions = BoolField("append_partitions", default=None)
     last_modified_time = Int64Field("last_modified_time", default=None)
     index_columns = ListField("index_columns", FieldTypes.string, default=None)
     index_dtypes = SeriesField("index_dtypes", default=None)
 
-    def __init__(self, memory_scale=None, **kw):
+    def __init__(self, memory_scale=None, dtype_backend=None, **kw):
         output_type = kw.pop("output_type", OutputType.dataframe)
         self._odps_entry = kw.pop("odps_entry", None)
+        dtype_backend = validate_dtype_backend(dtype_backend)
         super(DataFrameReadODPSTable, self).__init__(
-            memory_scale=memory_scale, _output_types=[output_type], **kw
+            memory_scale=memory_scale,
+            dtype_backend=dtype_backend,
+            _output_types=[output_type],
+            **kw,
         )
 
     @property
@@ -153,6 +162,7 @@ def read_odps_table(
     odps_entry: ODPS = None,
     string_as_binary: bool = None,
     append_partitions: bool = False,
+    dtype_backend: str = None,
     **kw,
 ):
     """
@@ -176,6 +186,8 @@ def read_odps_table(
     append_partitions: bool
         If True, will add all partition columns as selected columns when
         `columns` is not specified,
+    dtype_backend: {'numpy', 'pyarrow'}, default 'numpy'
+        Back-end data type applied to the resultant DataFrame (still experimental).
 
     Returns
     -------
@@ -202,9 +214,20 @@ def read_odps_table(
         else table.table_schema.simple_columns
     )
     table_columns = [c.name.lower() for c in cols]
-    table_dtypes = odps_schema_to_pandas_dtypes(
-        table.table_schema, with_partitions=True
+
+    if "use_arrow_dtype" in kw:
+        dtype_backend = dtype_backend or validate_dtype_backend(
+            kw.pop("use_arrow_dtype")
+        )
+    dtype_backend = validate_dtype_backend(
+        dtype_backend or options.dataframe.dtype_backend
     )
+
+    with option_context():
+        options.dataframe.dtype_backend = dtype_backend
+        table_dtypes = odps_schema_to_pandas_dtypes(
+            table.table_schema, with_partitions=True
+        )
     df_types = [table_dtypes[c] for c in table_columns]
 
     if isinstance(index_col, str):
@@ -246,7 +269,6 @@ def read_odps_table(
     dtypes = pd.Series(df_types, index=table_columns)
     chunk_bytes = kw.pop("chunk_bytes", None)
     chunk_size = kw.pop("chunk_size", None)
-    use_arrow_dtype = kw.pop("use_arrow_dtype", True)
 
     partitions = partitions or kw.get("partition")
     if isinstance(partitions, str):
@@ -261,7 +283,7 @@ def read_odps_table(
         partitions=partitions,
         dtypes=dtypes,
         columns=columns,
-        use_arrow_dtype=use_arrow_dtype,
+        dtype_backend=dtype_backend,
         string_as_binary=string_as_binary,
         append_partitions=append_partitions,
         last_modified_time=to_timestamp(table.last_data_modified_time),