"""``ParquetDataset`` loads/saves data from/to a Parquet file using an underlying
filesystem (e.g.: local, S3, GCS). It uses pandas to handle the Parquet file.
"""
from __future__ import annotations
import logging
from copy import deepcopy
from io import BytesIO
from pathlib import Path, PurePosixPath
from typing import Any
import fsspec
import pandas as pd
from kedro.io.core import (
PROTOCOL_DELIMITER,
AbstractVersionedDataset,
DatasetError,
Version,
get_filepath_str,
get_protocol_and_path,
)
from kedro_datasets._typing import TablePreview
logger = logging.getLogger(__name__)
[docs]
class ParquetDataset(AbstractVersionedDataset[pd.DataFrame, pd.DataFrame]):
"""``ParquetDataset`` loads/saves data from/to a Parquet file using an underlying
filesystem (e.g.: local, S3, GCS). It uses pandas to handle the Parquet file.
Example usage for the
`YAML API <https://kedro.readthedocs.io/en/stable/data/\
data_catalog_yaml_examples.html>`_:
.. code-block:: yaml
boats:
type: pandas.ParquetDataset
filepath: data/01_raw/boats.parquet
load_args:
engine: pyarrow
use_nullable_dtypes: True
save_args:
file_scheme: hive
has_nulls: False
engine: pyarrow
trucks:
type: pandas.ParquetDataset
filepath: abfs://container/02_intermediate/trucks.parquet
credentials: dev_abs
load_args:
columns: [name, gear, disp, wt]
index: name
save_args:
compression: GZIP
partition_on: [name]
Example usage for the
`Python API <https://kedro.readthedocs.io/en/stable/data/\
advanced_data_catalog_usage.html>`_:
.. code-block:: pycon
>>> from kedro_datasets.pandas import ParquetDataset
>>> import pandas as pd
>>>
>>> data = pd.DataFrame({"col1": [1, 2], "col2": [4, 5], "col3": [5, 6]})
>>>
>>> dataset = ParquetDataset(filepath=tmp_path / "test.parquet")
>>> dataset.save(data)
>>> reloaded = dataset.load()
>>> assert data.equals(reloaded)
"""
DEFAULT_LOAD_ARGS: dict[str, Any] = {}
DEFAULT_SAVE_ARGS: dict[str, Any] = {}
[docs]
def __init__( # noqa: PLR0913
self,
*,
filepath: str,
load_args: dict[str, Any] | None = None,
save_args: dict[str, Any] | None = None,
version: Version | None = None,
credentials: dict[str, Any] | None = None,
fs_args: dict[str, Any] | None = None,
metadata: dict[str, Any] | None = None,
) -> None:
"""Creates a new instance of ``ParquetDataset`` pointing to a concrete Parquet file
on a specific filesystem.
Args:
filepath: Filepath in POSIX format to a Parquet file prefixed with a protocol like
`s3://`. If prefix is not provided, `file` protocol (local filesystem) will be used.
The prefix should be any protocol supported by ``fsspec``.
It can also be a path to a directory. If the directory is
provided then it can be used for reading partitioned parquet files.
Note: `http(s)` doesn't support versioning.
load_args: Additional options for loading Parquet file(s).
Here you can find all available arguments when reading single file:
https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_parquet.html
Here you can find all available arguments when reading partitioned datasets:
https://arrow.apache.org/docs/python/generated/pyarrow.parquet.ParquetDataset.html#pyarrow.parquet.ParquetDataset.read
All defaults are preserved.
save_args: Additional saving options for saving Parquet file(s).
Here you can find all available arguments:
https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_parquet.html
All defaults are preserved. ``partition_cols`` is not supported.
version: If specified, should be an instance of ``kedro.io.core.Version``.
If its ``load`` attribute is None, the latest version will be loaded. If
its ``save`` attribute is None, save version will be autogenerated.
credentials: Credentials required to get access to the underlying filesystem.
E.g. for ``GCSFileSystem`` it should look like `{"token": None}`.
fs_args: Extra arguments to pass into underlying filesystem class constructor
(e.g. `{"project": "my-project"}` for ``GCSFileSystem``).
metadata: Any arbitrary metadata.
This is ignored by Kedro, but may be consumed by users or external plugins.
"""
_fs_args = deepcopy(fs_args) or {}
_credentials = deepcopy(credentials) or {}
protocol, path = get_protocol_and_path(filepath, version)
if protocol == "file":
_fs_args.setdefault("auto_mkdir", True)
self._protocol = protocol
self._storage_options = {**_credentials, **_fs_args}
self._fs = fsspec.filesystem(self._protocol, **self._storage_options)
self.metadata = metadata
super().__init__(
filepath=PurePosixPath(path),
version=version,
exists_function=self._fs.exists,
glob_function=self._fs.glob,
)
# Handle default load and save arguments
self._load_args = deepcopy(self.DEFAULT_LOAD_ARGS)
if load_args is not None:
self._load_args.update(load_args)
self._save_args = deepcopy(self.DEFAULT_SAVE_ARGS)
if save_args is not None:
self._save_args.update(save_args)
if "storage_options" in self._save_args or "storage_options" in self._load_args:
logger.warning(
"Dropping 'storage_options' for %s, "
"please specify them under 'fs_args' or 'credentials'.",
self._filepath,
)
self._save_args.pop("storage_options", None)
self._load_args.pop("storage_options", None)
def _describe(self) -> dict[str, Any]:
return {
"filepath": self._filepath,
"protocol": self._protocol,
"load_args": self._load_args,
"save_args": self._save_args,
"version": self._version,
}
def _load(self) -> pd.DataFrame:
load_path = str(self._get_load_path())
if self._protocol == "file":
# file:// protocol seems to misbehave on Windows
# (<urlopen error file not on local host>),
# so we don't join that back to the filepath;
# storage_options also don't work with local paths
return pd.read_parquet(load_path, **self._load_args)
load_path = f"{self._protocol}{PROTOCOL_DELIMITER}{load_path}"
return pd.read_parquet(
load_path, storage_options=self._storage_options, **self._load_args
)
def _save(self, data: pd.DataFrame) -> None:
save_path = get_filepath_str(self._get_save_path(), self._protocol)
if Path(save_path).is_dir():
raise DatasetError(
f"Saving {self.__class__.__name__} to a directory is not supported."
)
if "partition_cols" in self._save_args:
raise DatasetError(
f"{self.__class__.__name__} does not support save argument "
f"'partition_cols'. Please use 'kedro.io.PartitionedDataset' instead."
)
bytes_buffer = BytesIO()
data.to_parquet(bytes_buffer, **self._save_args)
with self._fs.open(save_path, mode="wb") as fs_file:
fs_file.write(bytes_buffer.getvalue())
self._invalidate_cache()
def _exists(self) -> bool:
try:
load_path = get_filepath_str(self._get_load_path(), self._protocol)
except DatasetError:
return False
return self._fs.exists(load_path)
def _release(self) -> None:
super()._release()
self._invalidate_cache()
def _invalidate_cache(self) -> None:
"""Invalidate underlying filesystem caches."""
filepath = get_filepath_str(self._filepath, self._protocol)
self._fs.invalidate_cache(filepath)
[docs]
def preview(self, nrows: int = 5) -> TablePreview:
"""
Generate a preview of the dataset with a specified number of rows.
Args:
nrows: The number of rows to include in the preview. Defaults to 5.
Returns:
dict: A dictionary containing the data in a split format.
"""
import pyarrow.parquet as pq
load_path = str(self._get_load_path())
table = pq.read_table(
load_path, columns=self._load_args.get("columns"), use_threads=True
)[:nrows]
data_preview = table.to_pandas()
return data_preview.to_dict(orient="split")
