urclib.fuzzylogic.combiner

Logic for combining crisp values from fuzzy logic analysis.

Module Contents

Classes

FLCombiner

Utility for applying defuzzification operators and combining
class urclib.fuzzylogic.combiner.FLCombiner(combinestr='', default_defuzz='centroid', **kwargs)

Bases: object

Utility for applying defuzzification operators and combining multiple decision spaces from fuzzy logic implications.

Parameters:

  • combineStr (str, optional) – Default python equation for combining defuzzed values.Defaults to empty str.

  • defaultDefuzz (str, optional) – String identifier of operator used to defuzz an implication by default. Options are: - centroid: Use the centroid of the decision space. This is the default. - bisector: Use the bisector value of the decision space. - smallest_of_maximum: The smallest x-value of all values that correspond to the maximum y-value. - largest_of_maximum: The largest x-value of all values that correspond to the maximum y-value. - mean_of_maximum: The mean of all x-values that correspond to points at the maximum y-value.

Keyword Arguments:

defuzzDict (dict) – Dictionary with a Fuzzy Logic set name as the key, and the defuzz operator to apply as the value. The valid defuzz values are identical to those listed for the defaultDefuzz argument.

property found_implication_names

A list of names of the implications expected to be passed in during evaluation.

Type:

list

set_defuzz_for_implication(impname, opname)

Set the defuzz operator to use for a specific implication.

Parameters:

  • impname (str) – The name of the fuzzy logic set to apply the defuzz operator to.

  • opname (str) – The key for the centroid operator. See the ‘defaultDefuzz’ in the class argument list.

parselogic(evalstr, refresh=False)

Parse the logic from instructions on how to combine defuzzed values.

Parameters:

  • evalstr (str) – The string of python logic to evaluate.

  • refresh (bool, optional) – If true, likely implication names will be extracted evalStr. This is useful for error checking. Defaults to False.

evaluate(implications, envargs=None, defuzzargs=None)

Evaluate the defuzzification operations and combining logic.

Parameters:

  • implications (dict) – Dictionary of implications to act upon.

  • envargs (dict,optional) – Additional arguments to pass to the execution environment.

  • defuzzargs (list,optional) – Additional arguments to pass on to defuzzification operations.

Returns:

A number value representing the combined defuzzed values.

Return type:

float

Raises:

fuzzylogic.FuzzyNoValError – If an expected implication is not found in implications, or if the implication processing encountered a fuzzy logic-specific error.

cleardefuzzforimplication(impl)

Clear the defuzzification operator applied to a specific implication.

Parameters:

impl (str) – The name of the implication to clear.

defuzzop_for_implication(impl, ret_default=False)

Retrieve the defuzzification operator for a specific implication.

Parameters:

  • impl (str) – The name of the implication to query.

  • ret_default (bool, optional) – Only effects implications assigned to the default operator; Return the default defuzz operator if True, or None if False. Defaults to False.

Returns:

The identifier of the defuzz operator of the implication. None: If the implication uses the default operator and retDefault is False.

Return type:

str

static parse_expected(evalstr)

Capture any likely variable names within a statement.

Parameters:

evalstr (str) – The statement to evaluate.

Returns:

list of strings representing suspected variable names in provided statement.

Return type:

list

static nodata_op(val, isval, isnotval=None)

Checks if a variable is a NO DATA value and returns the appropriate value.

Parameters:

  • val (float) – The value to evaluate for NO DATA.

  • isval (float) – The value to return if val is NO DATA.

  • isnotval (float, optional) – The value to return if val is not NO DATA

Returns:

static maxop(lhs, rhs, *args, **kwargs)

Overload of built-in max to properly handle the presence of a nodata sentinel.

Notes

The kwargs will be ignored, relying instead on our function.

Parameters:

  • lhs (object) – The first argument to check.

  • rhs (object) – The second argument to check.

  • *args – Additional values to compare.

  • **kwargs – ignored.

Returns:

The greater of the two objects, incorporating knowledge of the sentinels, if any present.

Return type:

object

static minop(lhs, rhs, *args, **kwargs)

Overload of built-in min to properly handle the presence of a nodata sentinel.

Notes

The kwargs will be ignored, relying instead on our function.

Parameters:

  • lhs (object) – The first argument to check.

  • rhs (object) – The second argument to check.

  • *args – Additional values to compare.

  • **kwargs – ignored.

Returns:

The lesser of the two objects, incorporating knowledge of the sentinels, if any present.

Return type:

object

static sumop(*args)

Standard sum operator.

Parameters:

*args – Values to sum together.

Returns:

The total of all combined values of args.

Return type:

number

static prodop(*args)

Standard product operator.

Parameters:

*args – The values to multiply together.

Returns:

The product of all combined values in args.

Return type:

number

static gammaop(gammaval, *args)

Gamma operator mimicing the Fuzzy Logic equivalent.

Parameters:

  • gammaval (number) – Value in the range of [0,1].

  • *args – The values to evaluate using the gamma value.

Returns:

The results of the gamma operation.

Raises:

ValueError – If gammaval is not in the range of [0,1].