types

This module gathers types used throughout the package.

BinaryScale

Type alias for binary scale

Function

Type alias for single argument python functions (callables).

alias of Callable[[Any], Any]

NormalScale

Type alias for normal scale

NumericFunction

Type alias for single argument python numeric functions (callables).

alias of Callable[[float], float]

class OrdinalScale(preference_direction=None, **kwargs)

Bases: Scale, ABC

This class defines a MCDA ordinal scale.

Parameters

preference_direction (Union[PreferenceDirection, None]) –

classmethod fit(data, preference_direction=None, **kwargs)

Create a scale that fits the data.

Parameters

• data (Series) –

• preference_direction (Optional[PreferenceDirection]) –

Returns

scale

Return type

Self

abstract property interval: Interval

Return interval on which scale is defined.

is_better(x, y)

Check if x is better than y according to this scale.

Parameters

• x (Any) –

• y (Any) –

Returns

Return type

bool

is_better_or_equal(x, y)

Check if x is better or equal to y according to this scale

Parameters

• x (Any) –

• y (Any) –

• scale –

Returns

Return type

bool

abstract label(x)

Return scale value corresponding to given number.

Parameters

x (float) –

Returns

Return type

Any

abstract property numeric: QuantitativeScale

Return corresponding numeric scale.

abstract range(nb=None)

Return range of value from scale.

Parameters

nb (Optional[int]) – number of values to return

Returns

Return type

List

abstract value(x)

Return corresponding numeric value.

Parameters

x (Any) –

Returns

Return type

float

class Relation(a, b)

Bases: ABC

This class represents a pairwise relation between two elements.

Parameters

• a (Any) – first element

• b (Any) – second element

Attribute a

Attribute b

Attribute DRAW_STYLE

(class) key args for plotting all instances

DRAW_STYLE: Dict[str, Any] = {'style': 'invis'}

compatible(other)

Check whether both relations can coexist in the same preference structure.

Relations are compatible if equal or having different elements pair.

Parameters

other (Relation) –

Returns

check result

Return type

bool

Warning

Does not check for relations’ validity!

property elements: Tuple[Any, Any]

Return elements of the relation

same_elements(relation)

Check whether the relations are about the same pair of alternatives.

Parameters

relation (Relation) – second relation

Returns

True if both relations share the same elements pair, False otherwise

Return type

bool

Warning

Does not check for relations’ validity!

static types()

Return list of relation types.

Returns

Return type

List

class Scale(**kwargs)

Bases: ABC

Basic abstract class for MCDA scale.

Attr preference_direction

classmethod fit(data, **kwargs)

Create a scale that fits the data.

Parameters

data (Series) –

Returns

scale

Return type

Self

is_better(x, y)

Check if x is better than y according to this scale.

Parameters

• x (Any) –

• y (Any) –

Returns

Raises

TypeError – if scale is not ordinal

Return type

bool

is_better_or_equal(x, y)

Check if x is better or equal to y according to this scale

Parameters

• x (Any) –

• y (Any) –

• scale –

Returns

Return type

bool

abstract range(nb=None)

Return range of value from scale.

Parameters

nb (Optional[int]) – number of values to return

Returns

Return type

List