pdstools.explanations.ExplanationsUtils¶

Classes¶

`_PREDICTOR_TYPE`	Create a collection of name/value pairs.
`_TABLE_NAME`	Create a collection of name/value pairs.
`_CONTRIBUTION_TYPE`	Create a collection of name/value pairs.
`_COL`	Create a collection of name/value pairs.
`_SPECIAL`	Create a collection of name/value pairs.
`_DEFAULT`	Create a collection of name/value pairs.
`ContextInfo`	dict() -> new empty dictionary
`ContextOperations`	Context related operations such as to filter unique contexts.

Module Contents¶

class _PREDICTOR_TYPE(*args, **kwds)¶

Bases: enum.Enum

Create a collection of name/value pairs.

Example enumeration:

>>> class Color(Enum):
...     RED = 1
...     BLUE = 2
...     GREEN = 3

Access them by:

attribute access:

>>> Color.RED
<Color.RED: 1>

value lookup:

>>> Color(1)
<Color.RED: 1>

name lookup:

>>> Color['RED']
<Color.RED: 1>

Enumerations can be iterated over, and know how many members they have:

>>> len(Color)
3

>>> list(Color)
[<Color.RED: 1>, <Color.BLUE: 2>, <Color.GREEN: 3>]

Methods can be added to enumerations, and members can have their own attributes – see the documentation for details.

NUMERIC = 'NUMERIC'¶

SYMBOLIC = 'SYMBOLIC'¶

class _TABLE_NAME(*args, **kwds)¶

Bases: enum.Enum

Create a collection of name/value pairs.

Example enumeration:

>>> class Color(Enum):
...     RED = 1
...     BLUE = 2
...     GREEN = 3

Access them by:

attribute access:

>>> Color.RED
<Color.RED: 1>

value lookup:

>>> Color(1)
<Color.RED: 1>

name lookup:

>>> Color['RED']
<Color.RED: 1>

Enumerations can be iterated over, and know how many members they have:

>>> len(Color)
3

>>> list(Color)
[<Color.RED: 1>, <Color.BLUE: 2>, <Color.GREEN: 3>]

Methods can be added to enumerations, and members can have their own attributes – see the documentation for details.

NUMERIC = 'numeric'¶

SYMBOLIC = 'symbolic'¶

NUMERIC_OVERALL = 'numeric_overall'¶

SYMBOLIC_OVERALL = 'symbolic_overall'¶

CREATE = 'create'¶

MODEL_CONTEXTS = 'model_contexts'¶

class _CONTRIBUTION_TYPE(default, alt, text)¶

Bases: enum.Enum

Create a collection of name/value pairs.

Example enumeration:

>>> class Color(Enum):
...     RED = 1
...     BLUE = 2
...     GREEN = 3

Access them by:

attribute access:

>>> Color.RED
<Color.RED: 1>

value lookup:

>>> Color(1)
<Color.RED: 1>

name lookup:

>>> Color['RED']
<Color.RED: 1>

Enumerations can be iterated over, and know how many members they have:

>>> len(Color)
3

>>> list(Color)
[<Color.RED: 1>, <Color.BLUE: 2>, <Color.GREEN: 3>]

Methods can be added to enumerations, and members can have their own attributes – see the documentation for details.

alt¶

text¶

classmethod validate_and_get_type(val)¶: Get the accepted contribution type which is validated against user input

CONTRIBUTION = ('contribution', 'contribution', 'average contribution')¶

CONTRIBUTION_ABS = ('contribution_abs', '|contribution|', 'absolute average contribution')¶

CONTRIBUTION_WEIGHTED = ('contribution_weighted', 'contribution weighted', 'weighted average contribution')¶

CONTRIBUTION_WEIGHTED_ABS = ('contribution_weighted_abs', '|contribution weighted|', 'absolute weighted average contribution')¶

FREQUENCY = ('frequency', 'frequency', 'frequency')¶

CONTRIBUTION_MIN = ('contribution_min', 'contribution min', 'minimum contribution')¶

CONTRIBUTION_MAX = ('contribution_max', 'contribution max', 'maximum contribution')¶

class _COL(*args, **kwds)¶

Bases: enum.Enum

Create a collection of name/value pairs.

Example enumeration:

>>> class Color(Enum):
...     RED = 1
...     BLUE = 2
...     GREEN = 3

Access them by:

attribute access:

>>> Color.RED
<Color.RED: 1>

value lookup:

>>> Color(1)
<Color.RED: 1>

name lookup:

>>> Color['RED']
<Color.RED: 1>

Enumerations can be iterated over, and know how many members they have:

>>> len(Color)
3

>>> list(Color)
[<Color.RED: 1>, <Color.BLUE: 2>, <Color.GREEN: 3>]

Methods can be added to enumerations, and members can have their own attributes – see the documentation for details.

PARTITON = 'partition'¶

PREDICTOR_NAME = 'predictor_name'¶

PREDICTOR_TYPE = 'predictor_type'¶

BIN_CONTENTS = 'bin_contents'¶

BIN_ORDER = 'bin_order'¶

CONTRIBUTION = 'contribution'¶

CONTRIBUTION_ABS = 'contribution_abs'¶

CONTRIBUTION_MIN = 'contribution_min'¶

CONTRIBUTION_MAX = 'contribution_max'¶

CONTRIBUTION_WEIGHTED = 'contribution_weighted'¶

CONTRIBUTION_WEIGHTED_ABS = 'contribution_weighted_abs'¶

FREQUENCY = 'frequency'¶

class _SPECIAL(*args, **kwds)¶

Bases: enum.Enum

Create a collection of name/value pairs.

Example enumeration:

>>> class Color(Enum):
...     RED = 1
...     BLUE = 2
...     GREEN = 3

Access them by:

attribute access:

>>> Color.RED
<Color.RED: 1>

value lookup:

>>> Color(1)
<Color.RED: 1>

name lookup:

>>> Color['RED']
<Color.RED: 1>

Enumerations can be iterated over, and know how many members they have:

>>> len(Color)
3

>>> list(Color)
[<Color.RED: 1>, <Color.BLUE: 2>, <Color.GREEN: 3>]

Methods can be added to enumerations, and members can have their own attributes – see the documentation for details.

REMAINING = 'remaining'¶

TOTAL_FREQUENCY = 'total_frequency'¶

MISSING = 'missing'¶

class _DEFAULT(*args, **kwds)¶

Bases: enum.Enum

Create a collection of name/value pairs.

Example enumeration:

>>> class Color(Enum):
...     RED = 1
...     BLUE = 2
...     GREEN = 3

Access them by:

attribute access:

>>> Color.RED
<Color.RED: 1>

value lookup:

>>> Color(1)
<Color.RED: 1>

name lookup:

>>> Color['RED']
<Color.RED: 1>

Enumerations can be iterated over, and know how many members they have:

>>> len(Color)
3

>>> list(Color)
[<Color.RED: 1>, <Color.BLUE: 2>, <Color.GREEN: 3>]

Methods can be added to enumerations, and members can have their own attributes – see the documentation for details.

TOP_N = 20¶

TOP_K = 20¶

DESCENDING = True¶

MISSING = True¶

REMAINING = True¶

class ContextInfo¶

Bases: TypedDict

dict() -> new empty dictionary dict(mapping) -> new dictionary initialized from a mapping object’s

(key, value) pairs

dict(iterable) -> new dictionary initialized as if via:: d = {} for k, v in iterable:

d[k] = v
dict(**kwargs) -> new dictionary initialized with the name=value pairs: in the keyword argument list. For example: dict(one=1, two=2)

context_key: str¶

context_value: str¶

class ContextOperations(aggregate: pdstools.explanations.Aggregate.Aggregate)¶

Bases: pdstools.utils.namespaces.LazyNamespace

Context related operations such as to filter unique contexts.

Parameters:

(Aggregate) (aggregate)
aggregate (pdstools.explanations.Aggregate.Aggregate)

aggregate(Aggregate)¶

Type:: The aggregate object.

_df(Optional[pl.DataFrame])¶

Type:: DataFrame containing context information.

_context_keys(Optional[list[str]])¶

Type:: list of context keys.

initialized(bool)¶

Type:: Flag indicating if the context operations have been initialized.

get_context_keys():: Returns the list of context keys from loaded data. Eg. [‘pyChannel’, ‘pyDirection’, …]

get_df(context_infos=None, with_partition_col=False):: Returns a DataFrame containing unique contexts If with_partition_col is True, includes the partition column. If context_infos is None, returns the full unique contexts, else filtered by the context Eg. with partition column: | pyChannel | pyDirection | … | partition | |-----------|————-|-----|———–| | channel1 | direction1 | … | {“partition”: {“pyChannel”: “channel1”, “pyDirection”: “direction1”}} | | channel1 | direction2 | … | {“partition”: {“pyChannel”: “channel1”, “pyDirection”: “direction2”}} |

get_list(context_infos=None, with_partition_col=False):

Returns a list[ContextInfo] containing unique contexts If with_partition_col is True, includes the partition column. If context_infos is None, returns the full unique contexts, else filtered by the context Eg. without partition column: [

{“pyChannel”: “channel1”, “pyDirection”: “direction1”, …}, {“pyChannel”: “channel1”, “pyDirection”: “direction2”, …},

]

get_context_info_str(context_info, sep="-"):: Returns a string representation of a single context information. Eg. channel1-direction1-…

dependencies = ['polars']¶

dependency_group = 'explanations'¶

aggregate¶

_df: polars.DataFrame | None = None¶

_context_keys: list[str] | None = None¶

initialized = False¶

_load()¶

get_context_keys()¶

get_df(context_infos: list[ContextInfo] | None = None, with_partition_col: bool = False) → polars.DataFrame¶

Get the DataFrame filtered by the provided context information.

Parameters:

context_infos (list[ContextInfo] | None)
with_partition_col (bool)

Return type:

polars.DataFrame

get_list(context_infos: list[ContextInfo] | None = None, with_partition_col: bool = False) → list[ContextInfo]¶

Get the list of context information filtered by the provided context information.

Parameters:

context_infos (list[ContextInfo] | None)
with_partition_col (bool)

Return type:

list[ContextInfo]

_filter_df_by_context_infos(df, context_infos)¶

static _get_filter_expression(context_infos)¶

static _get_clean_df(df: polars.DataFrame) → polars.DataFrame¶

Parameters:: df (polars.DataFrame)
Return type:: polars.DataFrame

static get_context_info_str(context_info: ContextInfo, sep: str = '-') → str¶

Parameters:

context_info (ContextInfo)
sep (str)

Return type:

str