import copy
import warnings
import weakref
from typing import Dict, Iterable, List, Set, Union
import pandas as pd
import woodwork.serialize as serialize
from woodwork.accessor_utils import (
_check_table_schema,
_is_dask_dataframe,
_is_dataframe,
_is_koalas_dataframe,
get_invalid_schema_message,
init_series,
)
from woodwork.exceptions import (
ColumnNotPresentError,
IndexTagRemovedWarning,
ParametersIgnoredWarning,
TypingInfoMismatchWarning,
)
from woodwork.indexers import _iLocIndexer, _locIndexer
from woodwork.logical_types import Datetime, LogicalType
from woodwork.statistics_utils import (
_get_describe_dict,
_get_mutual_information_dict,
_get_valid_mi_columns,
_get_value_counts,
_infer_temporal_frequencies,
)
from woodwork.table_schema import TableSchema
from woodwork.type_sys.utils import _is_numeric_series, col_is_datetime
from woodwork.typing import AnyDataFrame, ColumnName, UseStandardTagsDict
from woodwork.utils import (
_get_column_logical_type,
_parse_logical_type,
import_or_none,
import_or_raise,
)
dd = import_or_none("dask.dataframe")
ks = import_or_none("databricks.koalas")
[docs]class WoodworkTableAccessor:
[docs] def __init__(self, dataframe):
self._dataframe_weakref = weakref.ref(dataframe)
self._schema = None
[docs] def init(self, **kwargs):
"""Initializes Woodwork typing information for a DataFrame with a partial schema.
Logical type priority:
1. Types specified in ``logical_types``
2. Types specified in ``partial_schema``
3. Types inferred by ``ww.type_system.infer_logical_type``
Other Info priority:
1. Parameter passed in
2. Value specified in ``partial_schema``
Args:
schema (Woodwork.TableSchema, optional): Typing information to use for the DataFrame instead of performing inference.
Specified arguments will override the schema's typing information.
index (str, optional): Name of the index column.
time_index (str, optional): Name of the time index column.
logical_types (Dict[str -> LogicalType], optional): Dictionary mapping column names in
the DataFrame to the LogicalType for the column. Setting a column's logical type to None in this dict will
force a logical to be inferred.
already_sorted (bool, optional): Indicates whether the input DataFrame is already sorted on the time
index. If False, will sort the dataframe first on the time_index and then on the index (pandas DataFrame
only). Defaults to False.
name (str, optional): Name used to identify the DataFrame.
semantic_tags (dict, optional): Dictionary mapping column names in Woodwork to the
semantic tags for the column. The keys in the dictionary should be strings
that correspond to column names. There are two options for
specifying the dictionary values:
(str): If only one semantic tag is being set, a single string can be used as a value.
(list[str] or set[str]): If multiple tags are being set, a list or set of strings can be
used as the value.
Semantic tags will be set to an empty set for any column not included in the
dictionary.
table_metadata (Dict[str -> json serializable], optional): Dictionary containing extra metadata for Woodwork.
column_metadata (Dict[str -> Dict[str -> json serializable]], optional): Dictionary mapping column names
to that column's metadata dictionary.
use_standard_tags (bool, Dict[str -> bool], optional): Determines whether standard semantic tags will be
added to columns based on the specified logical type for the column.
If a single boolean is supplied, will apply the same use_standard_tags value to all columns.
A dictionary can be used to specify ``use_standard_tags`` values for individual columns.
Unspecified columns will use the default value of True.
column_descriptions (Dict[str -> str], optional): Dictionary mapping column names to column descriptions.
column_origins (str, Dict[str -> str], optional): Origin of each column. If a string is supplied, it is
used as the origin for all columns. A dictionary can be used to set origins for individual columns.
validate (bool, optional): Whether parameter and data validation should occur. Defaults to True. Warning:
Should be set to False only when parameters and data are known to be valid.
Any errors resulting from skipping validation with invalid inputs may not be easily understood.
"""
self.init_with_partial_schema(**kwargs)
[docs] def init_with_full_schema(
self, schema: TableSchema, validate: bool = True, **kwargs
) -> None:
"""Initializes Woodwork typing information for a DataFrame with a complete schema.
Args:
schema (Woodwork.TableSchema): Typing information to use for the DataFrame instead of performing inference. Note that any changes made to the
schema object after initialization will propagate to the DataFrame. Similarly, to avoid unintended typing information changes, the same schema
object should not be shared between DataFrames.
validate (bool, optional): Whether parameter and data validation should occur. Defaults to True. Warning: Should be set to False only when
parameters and data are known to be valid. Any errors resulting from skipping validation with invalid inputs may not be easily understood.
"""
if validate:
_check_schema(self._dataframe, schema)
_check_unique_column_names(self._dataframe)
self._schema = schema
extra_params = [key for key in kwargs]
if extra_params:
warnings.warn(
"A schema was provided and the following parameters were ignored: "
+ ", ".join(extra_params),
ParametersIgnoredWarning,
)
[docs] def init_with_partial_schema(
self,
schema: TableSchema = None,
index: str = None,
time_index: str = None,
logical_types: Dict[ColumnName, Union[str, LogicalType, None]] = None,
already_sorted: bool = False,
name: str = None,
semantic_tags: Dict[ColumnName, Union[str, List[str], Set[str]]] = None,
table_metadata: dict = None,
column_metadata: Dict[ColumnName, dict] = None,
use_standard_tags: Union[bool, UseStandardTagsDict] = None,
column_descriptions: Dict[ColumnName, str] = None,
column_origins: Union[str, Dict[ColumnName, str]] = None,
validate: bool = True,
**kwargs,
) -> None:
"""Initializes Woodwork typing information for a DataFrame with a partial schema.
Logical type priority:
1. Types specified in ``logical_types``
2. Types specified in ``partial_schema``
3. Types inferred by ``ww.type_system.infer_logical_type``
Other Info priority:
1. Parameter passed in
2. Value specified in ``partial_schema``
Args:
schema (Woodwork.TableSchema, optional): Typing information to use for the DataFrame instead of performing inference.
Specified arguments will override the schema's typing information.
index (str, optional): Name of the index column.
time_index (str, optional): Name of the time index column.
logical_types (Dict[str -> LogicalType], optional): Dictionary mapping column names in
the DataFrame to the LogicalType for the column. Setting a column's logical type to None in this dict will
force a logical to be inferred.
already_sorted (bool, optional): Indicates whether the input DataFrame is already sorted on the time
index. If False, will sort the dataframe first on the time_index and then on the index (pandas DataFrame
only). Defaults to False.
name (str, optional): Name used to identify the DataFrame.
semantic_tags (dict, optional): Dictionary mapping column names in Woodwork to the
semantic tags for the column. The keys in the dictionary should be strings
that correspond to column names. There are two options for
specifying the dictionary values:
(str): If only one semantic tag is being set, a single string can be used as a value.
(list[str] or set[str]): If multiple tags are being set, a list or set of strings can be
used as the value.
Semantic tags will be set to an empty set for any column not included in the
dictionary.
table_metadata (Dict[str -> json serializable], optional): Dictionary containing extra metadata for Woodwork.
column_metadata (Dict[str -> Dict[str -> json serializable]], optional): Dictionary mapping column names
to that column's metadata dictionary.
use_standard_tags (bool, Dict[str -> bool], optional): Determines whether standard semantic tags will be
added to columns based on the specified logical type for the column.
If a single boolean is supplied, will apply the same use_standard_tags value to all columns.
A dictionary can be used to specify ``use_standard_tags`` values for individual columns.
Unspecified columns will use the default value of True.
column_descriptions (Dict[str -> str], optional): Dictionary mapping column names to column descriptions.
column_origins (str, Dict[str -> str], optional): Origin of each column. If a string is supplied, it is
used as the origin for all columns. A dictionary can be used to set origins for individual columns.
validate (bool, optional): Whether parameter and data validation should occur. Defaults to True. Warning:
Should be set to False only when parameters and data are known to be valid.
Any errors resulting from skipping validation with invalid inputs may not be easily understood.
"""
if validate:
_validate_accessor_params(
self._dataframe,
index,
time_index,
logical_types,
schema,
use_standard_tags,
)
existing_logical_types = {}
existing_col_descriptions = {}
existing_col_metadata = {}
existing_use_standard_tags = {}
existing_semantic_tags = {}
existing_col_origins = {}
if schema: # pull schema parameters
name = name if name is not None else schema.name
index = index if index is not None else schema.index
time_index = time_index if time_index is not None else schema.time_index
table_metadata = table_metadata or schema.metadata
for col_name, col_schema in schema.columns.items():
existing_logical_types[col_name] = col_schema.logical_type
existing_semantic_tags[col_name] = (
col_schema.semantic_tags
- {"time_index"}
- {"index"}
- col_schema.logical_type.standard_tags
)
existing_col_descriptions[col_name] = col_schema.description
existing_col_origins[col_name] = col_schema.origin
existing_col_metadata[col_name] = col_schema.metadata
existing_use_standard_tags[col_name] = col_schema.use_standard_tags
# overwrite schema parameters with specified kwargs
logical_types = _infer_missing_logical_types(
self._dataframe, logical_types, existing_logical_types
)
column_descriptions = {
**existing_col_descriptions,
**(column_descriptions or {}),
}
column_metadata = {**existing_col_metadata, **(column_metadata or {})}
column_names = list(self._dataframe.columns)
use_standard_tags = _merge_use_standard_tags(
existing_use_standard_tags, use_standard_tags, column_names
)
semantic_tags = {**existing_semantic_tags, **(semantic_tags or {})}
column_origins = {**existing_col_origins, **(column_origins or {})}
self._schema = TableSchema(
column_names=column_names,
logical_types=logical_types,
name=name,
index=index,
time_index=time_index,
semantic_tags=copy.deepcopy(semantic_tags),
table_metadata=copy.deepcopy(table_metadata),
column_metadata=copy.deepcopy(column_metadata),
use_standard_tags=copy.deepcopy(use_standard_tags),
column_descriptions=column_descriptions,
column_origins=column_origins,
validate=validate,
**kwargs,
)
self._set_underlying_index()
if self._schema.time_index is not None:
self._sort_columns(already_sorted)
def __eq__(self, other, deep=True):
if not self._schema.__eq__(other.ww._schema, deep=deep):
return False
# Only check pandas DataFrames for equality
if (
deep
and isinstance(self._dataframe, pd.DataFrame)
and isinstance(other.ww._dataframe, pd.DataFrame)
):
return self._dataframe.equals(other.ww._dataframe)
return True
@_check_table_schema
def __getattr__(self, attr):
# Called if method is not present on the Accessor
# If the method is present on TableSchema, uses that method.
# If the method is present on DataFrame, uses that method.
if hasattr(self._schema, attr):
return self._make_schema_call(attr)
if hasattr(self._dataframe, attr):
return self._make_dataframe_call(attr)
else:
raise AttributeError(f"Woodwork has no attribute '{attr}'")
@_check_table_schema
def __getitem__(self, key):
if isinstance(key, list):
columns = set(self._dataframe.columns)
diff = list(set(key).difference(columns))
if diff:
raise ColumnNotPresentError(sorted(diff))
return self._get_subset_df_with_schema(key)
if key not in self._dataframe:
raise ColumnNotPresentError(key)
series = self._dataframe[key]
column = copy.deepcopy(self._schema.columns[key])
series.ww.init(schema=column, validate=False)
return series
def __setitem__(self, col_name, column):
series = tuple(pkg.Series for pkg in (pd, dd, ks) if pkg)
if not isinstance(column, series):
raise ValueError("New column must be of Series type")
if column.ww.schema is not None and "index" in column.ww.semantic_tags:
warnings.warn(
f'Cannot add "index" tag on {col_name} directly to the DataFrame. The "index" tag has been removed from {col_name}. To set this column as a Woodwork index, please use df.ww.set_index',
IndexTagRemovedWarning,
)
column.ww.set_semantic_tags(column.ww.semantic_tags - {"index"})
# Don't allow reassigning of index or time index with setitem
if self.index == col_name:
raise KeyError(
"Cannot reassign index. Change column name and then use df.ww.set_index to reassign index."
)
if self.time_index == col_name:
raise KeyError(
"Cannot reassign time index. Change column name and then use df.ww.set_time_index to reassign time index."
)
if column.ww._schema is None:
column = init_series(column, use_standard_tags=True)
self._dataframe[col_name] = column
self._schema.columns[col_name] = column.ww._schema
def __repr__(self):
"""A string representation of a Woodwork table containing typing information"""
return repr(self._get_typing_info())
def _repr_html_(self):
"""An HTML representation of a Woodwork table for IPython.display in Jupyter Notebooks
containing typing information and a preview of the data."""
return self._get_typing_info().to_html()
@_check_table_schema
def _get_typing_info(self):
"""Creates a DataFrame that contains the typing information for a Woodwork table."""
typing_info = self._schema._get_typing_info().copy()
typing_info.insert(0, "Physical Type", pd.Series(self.physical_types))
# Maintain the same column order used in the DataFrame
typing_info = typing_info.loc[list(self._dataframe.columns), :]
return typing_info
@property
@_check_table_schema
def name(self):
"""Name of the DataFrame"""
return self._schema.name
@name.setter
@_check_table_schema
def name(self, name):
"""Set name of the DataFrame"""
self._schema.name = name
@property
@_check_table_schema
def metadata(self):
"""Metadata of the DataFrame"""
return self._schema.metadata
@metadata.setter
@_check_table_schema
def metadata(self, metadata):
"""Set metadata of the DataFrame"""
self._schema.metadata = metadata
@property
def _dataframe(self):
return self._dataframe_weakref()
@property
@_check_table_schema
def iloc(self):
"""
Integer-location based indexing for selection by position.
``.iloc[]`` is primarily integer position based (from ``0`` to
``length-1`` of the axis), but may also be used with a boolean array.
If the selection result is a DataFrame or Series, Woodwork typing
information will be initialized for the returned object when possible.
Allowed inputs are:
An integer, e.g. ``5``.
A list or array of integers, e.g. ``[4, 3, 0]``.
A slice object with ints, e.g. ``1:7``.
A boolean array.
A ``callable`` function with one argument (the calling Series, DataFrame
or Panel) and that returns valid output for indexing (one of the above).
This is useful in method chains, when you don't have a reference to the
calling object, but would like to base your selection on some value.
"""
return _iLocIndexer(self._dataframe)
@property
@_check_table_schema
def loc(self):
"""
Access a group of rows by label(s) or a boolean array.
``.loc[]`` is primarily label based, but may also be used with a
boolean array.
If the selection result is a DataFrame or Series, Woodwork typing
information will be initialized for the returned object when possible.
Allowed inputs are:
A single label, e.g. ``5`` or ``'a'``, (note that ``5`` is
interpreted as a *label* of the index, and **never** as an
integer position along the index).
A list or array of labels, e.g. ``['a', 'b', 'c']``.
A slice object with labels, e.g. ``'a':'f'``.
A boolean array of the same length as the axis being sliced,
e.g. ``[True, False, True]``.
An alignable boolean Series. The index of the key will be aligned before
masking.
An alignable Index. The Index of the returned selection will be the input.
A ``callable`` function with one argument (the calling Series or
DataFrame) and that returns valid output for indexing (one of the above)
"""
return _locIndexer(self._dataframe)
@property
def schema(self):
"""A copy of the Woodwork typing information for the DataFrame."""
if self._schema:
return copy.deepcopy(self._schema)
@property
@_check_table_schema
def physical_types(self):
"""A dictionary containing physical types for each column"""
return {
col_name: self._schema.logical_types[col_name]._get_valid_dtype(
type(self._dataframe[col_name])
)
for col_name in self._dataframe.columns
}
@property
@_check_table_schema
def types(self):
"""DataFrame containing the physical dtypes, logical types and semantic
tags for the schema."""
return self._get_typing_info()
@property
@_check_table_schema
def logical_types(self):
"""A dictionary containing logical types for each column"""
return self._schema.logical_types
@property
@_check_table_schema
def semantic_tags(self):
"""A dictionary containing semantic tags for each column"""
return self._schema.semantic_tags
@property
@_check_table_schema
def index(self):
"""The index column for the table"""
return self._schema.index
@property
@_check_table_schema
def time_index(self):
"""The time index column for the table"""
return self._schema.time_index
@property
@_check_table_schema
def use_standard_tags(self):
"""A dictionary containing the use_standard_tags setting for each column in the table"""
return self._schema.use_standard_tags
[docs] @_check_table_schema
def set_index(self, new_index):
"""Sets the index column of the DataFrame. Adds the 'index' semantic tag to the column
and clears the tag from any previously set index column.
Setting a column as the index column will also cause any previously set standard
tags for the column to be removed.
Clears the DataFrame's index by passing in None.
Args:
new_index (str): The name of the column to set as the index
"""
self._schema.set_index(new_index)
if self._schema.index is not None:
_check_index(self._dataframe, self._schema.index)
self._set_underlying_index()
[docs] @_check_table_schema
def set_time_index(self, new_time_index):
"""Set the time index. Adds the 'time_index' semantic tag to the column and
clears the tag from any previously set index column
Args:
new_time_index (str): The name of the column to set as the time index.
If None, will remove the time_index.
"""
self._schema.set_time_index(new_time_index)
[docs] @_check_table_schema
def set_types(self, logical_types=None, semantic_tags=None, retain_index_tags=True):
"""Update the logical type and semantic tags for any columns names in the provided types dictionaries,
updating the Woodwork typing information for the DataFrame.
Args:
logical_types (Dict[str -> str], optional): A dictionary defining the new logical types for the
specified columns.
semantic_tags (Dict[str -> str/list/set], optional): A dictionary defining the new semantic_tags for the
specified columns.
retain_index_tags (bool, optional): If True, will retain any index or time_index
semantic tags set on the column. If False, will replace all semantic tags any time a column's
semantic tags or logical type changes. Defaults to True.
"""
logical_types = logical_types or {}
logical_types = {
col_name: _parse_logical_type(ltype, col_name)
for col_name, ltype in logical_types.items()
}
self._schema.set_types(
logical_types=logical_types,
semantic_tags=semantic_tags,
retain_index_tags=retain_index_tags,
)
# go through changed ltypes and update dtype if necessary
for col_name, logical_type in logical_types.items():
series = self._dataframe[col_name]
updated_series = logical_type.transform(series)
if updated_series is not series:
self._dataframe[col_name] = updated_series
[docs] @_check_table_schema
def select(self, include=None, exclude=None, return_schema=False):
"""Create a DataFrame with Woodwork typing information initialized
that includes only columns whose Logical Type and semantic tags match
conditions specified in the list of types and tags to include or exclude.
Values for both ``include`` and ``exclude`` cannot be provided in a
single call.
If no matching columns are found, an empty DataFrame will be returned.
Args:
include (str or LogicalType or list[str or LogicalType]): Logical
types, semantic tags to include in the DataFrame.
exclude (str or LogicalType or list[str or LogicalType]): Logical
types, semantic tags to exclude from the DataFrame.
return_schema (boolen): If True, return only the schema for the
matching columns. Defaults to False
Returns:
DataFrame: The subset of the original DataFrame that matches the
conditions specified by ``include`` or ``exclude``. Has Woodwork
typing information initialized.
"""
if include is not None and exclude is not None:
raise ValueError(
"Cannot specify values for both 'include' and 'exclude' in a single call."
)
if include is None and exclude is None:
raise ValueError("Must specify values for either 'include' or 'exclude'.")
cols_to_include = self._schema._filter_cols(include, exclude)
if return_schema:
return self._schema.get_subset_schema(cols_to_include)
return self._get_subset_df_with_schema(cols_to_include)
[docs] @_check_table_schema
def to_dictionary(self):
"""Get a dictionary representation of the Woodwork typing information.
Returns:
dict: Description of the typing information.
"""
return serialize.typing_info_to_dict(self._dataframe)
[docs] @_check_table_schema
def to_disk(
self, path, format="csv", compression=None, profile_name=None, **kwargs
):
"""Write Woodwork table to disk in the format specified by `format`, location specified by `path`.
Path could be a local path or an S3 path.
If writing to S3 a tar archive of files will be written.
Note:
As the engine `fastparquet` cannot handle nullable pandas dtypes, `pyarrow` will be used
for serialization to parquet.
Args:
path (str) : Location on disk to write to (will be created as a directory)
format (str) : Format to use for writing Woodwork data. Defaults to csv. Possible values are: {'csv', 'pickle', 'parquet'}.
compression (str) : Name of the compression to use. Possible values are: {'gzip', 'bz2', 'zip', 'xz', None}.
profile_name (str) : Name of AWS profile to use, False to use an anonymous profile, or None.
kwargs (keywords) : Additional keyword arguments to pass as keywords arguments to the underlying serialization method or to specify AWS profile.
"""
if format == "csv":
default_csv_kwargs = {
"sep": ",",
"encoding": "utf-8",
"engine": "python",
"index": False,
}
if _is_koalas_dataframe(self._dataframe):
default_csv_kwargs["multiline"] = True
default_csv_kwargs["ignoreLeadingWhitespace"] = False
default_csv_kwargs["ignoreTrailingWhitespace"] = False
kwargs = {**default_csv_kwargs, **kwargs}
elif format in ["parquet", "orc"]:
import_error_message = (
f"The pyarrow library is required to serialize to {format}.\n"
"Install via pip:\n"
" pip install pyarrow\n"
"Install via conda:\n"
" conda install pyarrow -c conda-forge"
)
import_or_raise("pyarrow", import_error_message)
kwargs["engine"] = "pyarrow"
serialize.write_woodwork_table(
self._dataframe,
path,
format=format,
compression=compression,
profile_name=profile_name,
**kwargs,
)
def _sort_columns(self, already_sorted):
if _is_dask_dataframe(self._dataframe) or _is_koalas_dataframe(self._dataframe):
already_sorted = True # Skip sorting for Dask and Koalas input
if not already_sorted:
sort_cols = [self._schema.time_index, self._schema.index]
if self._schema.index is None:
sort_cols = [self._schema.time_index]
self._dataframe.sort_values(sort_cols, inplace=True)
def _set_underlying_index(self):
"""Sets the index of the underlying DataFrame to match the index column as
specified by the TableSchema. Does not change the underlying index if no Woodwork index is
specified. Only sets underlying index for pandas DataFrames.
"""
if isinstance(self._dataframe, pd.DataFrame) and self._schema.index is not None:
self._dataframe.set_index(self._schema.index, drop=False, inplace=True)
# Drop index name to not overlap with the original column
self._dataframe.index.name = None
def _make_schema_call(self, attr):
"""Forwards the requested attribute onto the schema object.
Results are that of the Woodwork.TableSchema class."""
schema_attr = getattr(self._schema, attr)
if callable(schema_attr):
def wrapper(*args, **kwargs):
return schema_attr(*args, **kwargs)
return wrapper
return schema_attr
def _make_dataframe_call(self, attr):
"""Forwards the requested attribute onto the dataframe object. Intercepts return value,
attempting to initialize Woodwork with the current schema when a new DataFrame is returned.
Confirms schema is still valid for the original DataFrame."""
dataframe_attr = getattr(self._dataframe, attr)
if callable(dataframe_attr):
def wrapper(*args, **kwargs):
# Make DataFrame call and intercept the result
result = dataframe_attr(*args, **kwargs)
# Try to initialize Woodwork with the existing schema
if _is_dataframe(result):
invalid_schema_message = get_invalid_schema_message(
result, self._schema
)
if invalid_schema_message:
warnings.warn(
TypingInfoMismatchWarning().get_warning_message(
attr, invalid_schema_message, "DataFrame"
),
TypingInfoMismatchWarning,
)
else:
copied_schema = self.schema
result.ww.init_with_full_schema(
schema=copied_schema, validate=False
)
else:
# Confirm that the schema is still valid on original DataFrame
# Important for inplace operations
invalid_schema_message = get_invalid_schema_message(
self._dataframe, self._schema
)
if invalid_schema_message:
warnings.warn(
TypingInfoMismatchWarning().get_warning_message(
attr, invalid_schema_message, "DataFrame"
),
TypingInfoMismatchWarning,
)
self._schema = None
# Always return the results of the DataFrame operation whether or not Woodwork is initialized
return result
return wrapper
# Directly return non-callable DataFrame attributes
return dataframe_attr
def _get_subset_df_with_schema(self, cols_to_include, inplace=False):
"""Creates a new DataFrame from a list of column names with Woodwork initialized,
retaining all typing information and maintaining the DataFrame's column order."""
if inplace:
if _is_dask_dataframe(self._dataframe):
raise ValueError("Drop inplace not supported for Dask")
if _is_koalas_dataframe(self._dataframe):
raise ValueError("Drop inplace not supported for Koalas")
assert all([col_name in self._schema.columns for col_name in cols_to_include])
new_schema = self._schema.get_subset_schema(cols_to_include)
if inplace:
cols_to_drop = [
col_name
for col_name in self._dataframe.columns
if col_name not in cols_to_include
]
self._dataframe.drop(cols_to_drop, axis="columns", inplace=True)
self.init_with_full_schema(schema=new_schema, validate=False)
return
new_df = self._dataframe[cols_to_include]
new_df.ww.init_with_full_schema(schema=new_schema, validate=False)
return new_df
[docs] @_check_table_schema
def pop(self, column_name):
"""Return a Series with Woodwork typing information and remove it from the DataFrame.
Args:
column (str): Name of the column to pop.
Returns:
Series: Popped series with Woodwork initialized
"""
if column_name not in self._dataframe.columns:
raise ColumnNotPresentError(column_name)
series = self._dataframe.pop(column_name)
# Initialize Woodwork typing info for series
series.ww.init(schema=self.schema.columns[column_name], validate=False)
# Update schema to not include popped column
del self._schema.columns[column_name]
return series
[docs] @_check_table_schema
def drop(self, columns, inplace=False):
"""Drop specified columns from a DataFrame.
Args:
columns (str or list[str]): Column name or names to drop. Must be present in the DataFrame.
inplace (bool): If False, return a copy. Otherwise, do operation inplace and return None.
Returns:
DataFrame or None: DataFrame with the specified columns removed, maintaining Woodwork typing information or None if inplace=True. Only possible for pandas dataframes.
Note:
This method is used for removing columns only. To remove rows with ``drop``, go through the
DataFrame directly and then reinitialize Woodwork with ``DataFrame.ww.init``
instead of calling ``DataFrame.ww.drop``.
"""
if not isinstance(columns, (list, set)):
columns = [columns]
not_present = [col for col in columns if col not in self._dataframe.columns]
if not_present:
raise ColumnNotPresentError(not_present)
return self._get_subset_df_with_schema(
[col for col in self._dataframe.columns if col not in columns],
inplace=inplace,
)
[docs] @_check_table_schema
def rename(self, columns, inplace=False):
"""Renames columns in a DataFrame, maintaining Woodwork typing information.
Args:
columns (Dict[str -> str]): A dictionary mapping current column names to new column names.
inplace (bool): If False, return a copy. Otherwise, do operation inplace and return None.
Returns:
DataFrame or None: DataFrame with the specified columns renamed, maintaining Woodwork typing information or None if inplace=True. Only possible for pandas dataframes.
"""
new_schema = self._schema.rename(columns)
if inplace:
if _is_dask_dataframe(self._dataframe):
raise ValueError("Rename inplace not supported for Dask")
if _is_koalas_dataframe(self._dataframe):
raise ValueError("Rename inplace not supported for Koalas")
self._dataframe.rename(columns=columns, inplace=True)
self.init_with_full_schema(schema=new_schema)
return
new_df = self._dataframe.rename(columns=columns)
new_df.ww.init_with_full_schema(schema=new_schema, validate=False)
return new_df
def get_valid_mi_columns(self, include_index=False):
"""Retrieves a list of columns from the DataFrame with valid Logical Types that support mutual
information. Use get_valid_mi_types to see which Logical Types support mutual information:
>>> from woodwork.utils import get_valid_mi_types
>>> get_valid_mi_types()
[Age, AgeFractional, AgeNullable, Boolean, BooleanNullable, Categorical, CountryCode, Datetime, Double, Integer, IntegerNullable, Ordinal, PostalCode, SubRegionCode]
Args:
include_index (bool): If True, the column specified as the index will be
included as long as its LogicalType is valid for mutual information calculations.
If False, the index column will not have mutual information calculated for it.
Defaults to False.
Returns:
list: A list of column names that have valid Logical Types that support mutual information.
"""
return _get_valid_mi_columns(self._dataframe, include_index)
[docs] @_check_table_schema
def describe_dict(
self,
include=None,
callback=None,
extra_stats=False,
bins=10,
top_x=10,
recent_x=10,
):
"""Calculates statistics for data contained in the DataFrame.
Args:
include (list[str or LogicalType], optional): filter for what columns to include in the
statistics returned. Can be a list of column names, semantic tags, logical types, or a list
combining any of the three. It follows the most broad specification. Favors logical types
then semantic tag then column name. If no matching columns are found, an empty DataFrame
will be returned.
callback (callable, optional): function to be called with incremental updates. Has the following parameters:
- update (int): change in progress since last call
- progress (int): the progress so far in the calculations
- total (int): the total number of calculations to do
- unit (str): unit of measurement for progress/total
- time_elapsed (float): total time in seconds elapsed since start of call
extra_stats (bool): If True, will calculate a histogram for numeric columns, top values
for categorical columns and value counts for the most recent values in datetime columns. Will also
calculate value counts within the range of values present for integer columns if the range of
values present is less than or equal to than the number of bins used to compute the histogram.
Output can be controlled by bins, top_x and recent_x parameters.
bins (int): Number of bins to use when calculating histogram for numeric columns. Defaults to 10.
Will be ignored unless extra_stats=True.
top_x (int): Number of items to return when getting the most frequently occurring values for categorical
columns. Defaults to 10. Will be ignored unless extra_stats=True.
recent_x (int): Number of values to return when calculating value counts for the most recent dates in
datetime columns. Defaults to 10. Will be ignored unless extra_stats=True.
Returns:
Dict[str -> dict]: A dictionary with a key for each column in the data or for each column
matching the logical types, semantic tags or column names specified in ``include``, paired
with a value containing a dictionary containing relevant statistics for that column.
"""
return _get_describe_dict(
self._dataframe,
include=include,
callback=callback,
extra_stats=extra_stats,
bins=bins,
top_x=top_x,
recent_x=recent_x,
)
[docs] def describe(self, include=None, callback=None):
"""Calculates statistics for data contained in the DataFrame.
Args:
include (list[str or LogicalType], optional): filter for what columns to include in the
statistics returned. Can be a list of column names, semantic tags, logical types, or a list
combining any of the three. It follows the most broad specification. Favors logical types
then semantic tag then column name. If no matching columns are found, an empty DataFrame
will be returned.
callback (callable, optional): function to be called with incremental updates. Has the following parameters:
- update (int): change in progress since last call
- progress (int): the progress so far in the calculations
- total (int): the total number of calculations to do
- unit (str): unit of measurement for progress/total
- time_elapsed (float): total time in seconds elapsed since start of call
Returns:
pd.DataFrame: A Dataframe containing statistics for the data or the subset of the original
DataFrame that contains the logical types, semantic tags, or column names specified
in ``include``.
"""
results = self.describe_dict(include=include, callback=callback)
index_order = [
"physical_type",
"logical_type",
"semantic_tags",
"count",
"nunique",
"nan_count",
"mean",
"mode",
"std",
"min",
"first_quartile",
"second_quartile",
"third_quartile",
"max",
"num_true",
"num_false",
]
return pd.DataFrame(results).reindex(index_order)
[docs] @_check_table_schema
def value_counts(self, ascending=False, top_n=10, dropna=False):
"""Returns a list of dictionaries with counts for the most frequent values in each column (only
for columns with `category` as a standard tag).
Args:
ascending (bool): Defines whether each list of values should be sorted most frequent
to least frequent value (False), or least frequent to most frequent value (True).
Defaults to False.
top_n (int): the number of top values to retrieve. Defaults to 10.
dropna (bool): determines whether to remove NaN values when finding frequency. Defaults
to False.
Returns:
list(dict): a list of dictionaries for each categorical column with keys `count`
and `value`.
"""
return _get_value_counts(self._dataframe, ascending, top_n, dropna)
[docs] @_check_table_schema
def infer_temporal_frequencies(self, temporal_columns=None):
"""Infers the observation frequency (daily, biweekly, yearly, etc) of each temporal column
in the DataFrame. Temporal columns are ones with the logical type Datetime or Timedelta.
Not supported for Dask and Koalas DataFrames.
Args:
temporal_columns (list[str], optional): Columns for which frequencies should be inferred. Must be columns
that are present in the DataFrame and are temporal in nature. Defaults to None. If not
specified, all temporal columns will have their frequencies inferred.
Returns:
(dict): A dictionary where each key is a temporal column from the DataFrame, and the
value is its observation frequency represented as a pandas offset alias string (D, M, Y, etc.)
or None if no uniform frequency was present in the data.
Note:
The pandas util ``pd.infer_freq``, which is used in this method, has the following behaviors:
- If even one row in a column does not follow the frequency seen in the remaining rows,
no frequency will be inferred. Example of otherwise daily data that skips one day:
``['2011-01-03', '2011-01-04', '2011-01-05', '2011-01-07']``.
- If any NaNs are present in the data, no frequency will be inferred.
- Pandas will use the largest offset alias available to it, so ``W`` will be inferred for weekly data instead of ``7D``.
The list of available offset aliases, which include aliases such as ``B`` for business day or ``N`` for nanosecond,
can be found at https://pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases
- Offset aliases can be combined to create something like ``2d1H``, which could also be expressed as '49H'.
Pandas' frequency inference will return the lower common alias, ``49H``, in situations when it'd otherwise
need to combine aliases.
- Offset strings can contain more information than just the offset alias. For example, a date range
``pd.date_range(start="2020-01-01", freq="w", periods=10)`` will be inferred to have frequency ``W-SUN``.
That string is an offset alias with an anchoring suffix that indicates that the data is not only
observed at a weekly frequency, but that all the dates are on Sundays. More anchored offsets
can be seen here: https://pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#anchored-offsets
- Some frequencies that can be defined for a ``pd.date_range`` cannot then be re-inferred by pandas' ``pd.infer_freq``.
One example of this can be seen when using the business day offset alias ``B``
``pd.date_range(start="2020-01-01", freq="4b", periods=10)``, which is a valid ``freq``
parameter when building the date range, but is not then inferrable.
"""
return _infer_temporal_frequencies(
self._dataframe, temporal_columns=temporal_columns
)
def _validate_accessor_params(
dataframe, index, time_index, logical_types, schema, use_standard_tags
) -> None:
_check_unique_column_names(dataframe)
if use_standard_tags is not None:
_check_use_standard_tags(use_standard_tags)
if schema is not None:
_check_partial_schema(dataframe, schema)
if index is None:
# if no index was passed in as a parameter we need to validate the existing index
index = schema.index
if index is not None:
_check_index(dataframe, index)
if logical_types:
_check_logical_types(dataframe.columns, logical_types)
if time_index is not None:
datetime_format = None
logical_type = None
if logical_types is not None and time_index in logical_types:
logical_type = logical_types[time_index]
if type(logical_types[time_index]) == Datetime:
datetime_format = logical_types[time_index].datetime_format
_check_time_index(
dataframe,
time_index,
datetime_format=datetime_format,
logical_type=logical_type,
)
def _check_unique_column_names(dataframe):
if not dataframe.columns.is_unique:
raise IndexError("Dataframe cannot contain duplicate columns names")
def _check_index(dataframe, index):
if index not in dataframe.columns:
# User specifies an index that is not in the dataframe
raise ColumnNotPresentError(
f"Specified index column `{index}` not found in dataframe"
)
if (
index is not None
and isinstance(dataframe, pd.DataFrame)
and not dataframe[index].is_unique
):
# User specifies an index that is in the dataframe but not unique
# Does not check for Dask as Dask does not support is_unique
raise IndexError("Index column must be unique")
def _check_time_index(dataframe, time_index, datetime_format=None, logical_type=None):
if time_index not in dataframe.columns:
raise ColumnNotPresentError(
f"Specified time index column `{time_index}` not found in dataframe"
)
if not (
_is_numeric_series(dataframe[time_index], logical_type)
or col_is_datetime(dataframe[time_index], datetime_format=datetime_format)
):
raise TypeError("Time index column must contain datetime or numeric values")
def _check_logical_types(dataframe_columns, logical_types):
if not isinstance(logical_types, dict):
raise TypeError("logical_types must be a dictionary")
cols_not_found = set(logical_types.keys()).difference(set(dataframe_columns))
if cols_not_found:
raise ColumnNotPresentError(
"logical_types contains columns that are not present in "
f"dataframe: {sorted(list(cols_not_found))}"
)
def _check_schema(dataframe, schema):
if not isinstance(schema, TableSchema):
raise TypeError("Provided schema must be a Woodwork.TableSchema object.")
invalid_schema_message = get_invalid_schema_message(dataframe, schema)
if invalid_schema_message:
raise ValueError(
f"Woodwork typing information is not valid for this DataFrame: {invalid_schema_message}"
)
def _check_partial_schema(dataframe, schema: TableSchema) -> None:
if not isinstance(schema, TableSchema):
raise TypeError("Provided schema must be a Woodwork.TableSchema object.")
dataframe_cols = set(dataframe.columns)
schema_cols = set(schema.columns.keys())
schema_cols_not_in_df = schema_cols - dataframe_cols
if schema_cols_not_in_df:
raise ColumnNotPresentError(
f"The following columns in the typing information were missing from the DataFrame: "
f"{schema_cols_not_in_df}"
)
def _check_use_standard_tags(use_standard_tags):
if not isinstance(use_standard_tags, (bool, dict)):
raise TypeError("use_standard_tags must be a dictionary or a boolean")
def _infer_missing_logical_types(
dataframe: AnyDataFrame,
force_logical_types: Dict[ColumnName, Union[str, LogicalType]] = None,
existing_logical_types: Dict[ColumnName, Union[str, LogicalType]] = None,
):
"""Performs type inference and updates underlying data"""
force_logical_types = force_logical_types or {}
existing_logical_types = existing_logical_types or {}
parsed_logical_types = {}
for name in dataframe.columns:
series = dataframe[name]
logical_type = (
force_logical_types.get(name)
if name in force_logical_types
else existing_logical_types.get(name)
)
parsed_logical_types[name] = _get_column_logical_type(
series, logical_type, name
)
updated_series = parsed_logical_types[name].transform(series)
if updated_series is not series:
dataframe[name] = updated_series
return parsed_logical_types
def _merge_use_standard_tags(
existing_use_standard_tags: UseStandardTagsDict = None,
use_standard_tags: Union[bool, UseStandardTagsDict] = None,
column_names: Iterable[ColumnName] = None,
default_use_standard_tag: bool = True,
) -> UseStandardTagsDict:
"""Combines existing and kwarg use_standard_tags and returns a UseStandardTagsDict with all column names
Priority when merging:
1. use_standard tags
2. existing_use_standard_tags
3. default_use_standard_tag
"""
column_names = column_names or []
if isinstance(use_standard_tags, bool):
use_standard_tags = {col_name: use_standard_tags for col_name in column_names}
else:
use_standard_tags = {
**{col_name: default_use_standard_tag for col_name in column_names},
**(existing_use_standard_tags or {}),
**(use_standard_tags or {}),
}
return use_standard_tags
@pd.api.extensions.register_dataframe_accessor("ww")
class PandasTableAccessor(WoodworkTableAccessor):
pass
if dd:
@dd.extensions.register_dataframe_accessor("ww")
class DaskTableAccessor(WoodworkTableAccessor):
pass
if ks:
from databricks.koalas.extensions import register_dataframe_accessor
@register_dataframe_accessor("ww")
class KoalasTableAccessor(WoodworkTableAccessor):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
if not ks.get_option("compute.ops_on_diff_frames"):
ks.set_option("compute.ops_on_diff_frames", True)