Temporal fuzzy sets

Expansion of the base fuzzy sets, adding temporal fuzzy sets.

Contains functions to model the temporal fuzzy sets, temporal fuzzy variables and temporal rule bases. It also contains functions to evaluate the fuzzy rulebases obtained.

ex_fuzzy.temporal.NEW_FUZZY_SETS

alias of FUZZY_SETS

ex_fuzzy.temporal.TMP_FUZZY_SETS

alias of NEW_FUZZY_SETS

class ex_fuzzy.temporal.TemporalFuzzyRulesClassifier(nRules: int = 30, nAnts: int = 4, fuzzy_type: FUZZY_SETS = None, tolerance: float = 0.0, n_linguistic_variables: int = 0, verbose=False, linguistic_variables: list[fuzzyVariable] = None, domain: list[float] = None, n_class: int = None, precomputed_rules: MasterRuleBase = None, runner: int = 1)

Bases: BaseFuzzyRulesClassifier

Class that is used as a classifier for a fuzzy rule based system with time dependencies. Supports precomputed and optimization of the linguistic variables.

fit(X: array, y: array, n_gen: int = 50, pop_size: int = 10, time_moments: array = None, checkpoints: int = 0)

Fits a fuzzy rule based classifier using a genetic algorithm to the given data.

Parameters:

• X – numpy array samples x features

• y – labels. integer array samples (x 1)

• n_gen – integer. Number of generations to run the genetic algorithm.

• pop_size – integer. Population size for each gneration.

• time_moments – array of integers. Time moments associated to each sample (when temporal dependencies are present)

Returns:

None. The classifier is fitted to the data.

forward(X: array, time_moments: list[int] = None) → array

Returns the predicted class for each sample.

Parameters:

• X – np array samples x features.

• time_moments – list of integers. Time moments associated to each sample (when temporal dependencies are present)

Returns:

np array samples (x 1) with the predicted class.

get_rulebase() → list[array]

Get the rulebase obtained after fitting the classifier to the data.

Returns:

a matrix format for the rulebase.

plot_fuzzy_variables() → None

Plot the fuzzy partitions in each fuzzy variable.

ex_fuzzy.temporal.eval_temporal_fuzzy_model(fl_classifier: BaseFuzzyRulesClassifier, X_train: array, y_train: array, X_test: array, y_test: array, time_moments: list[int] = None, test_time_moments: list[int] = None, plot_rules=True, print_rules=True, plot_partitions=True, return_rules=False, print_accuracy=True, print_matthew=True) → None

Function that evaluates a fuzzy rule based model. It also plots the rules and the fuzzy partitions.

Parameters:

• fl_classifier – Fuzzy rule based model.

• X_train – Training data.

• y_train – Training labels.

• X_test – Test data.

• y_test – Test labels.

• plot_rules – If True, it plots the rules.

• print_rules – If True, it prints the rules.

• plot_partitions – If True, it plots the fuzzy partitions.

Returns:

None

class ex_fuzzy.temporal.temporalFS(std_fuzzy_set: FS, conditional_variable: array)

Bases: FS

Class to implement temporal fuzzy sets.

fix_time(time: int) → None

Fixes the time of the temporal fuzzy set.

Parameters:

time – int. Time moment to fix.

Returns:

FS. Fuzzy set with the fixed time.

inside_type() → FUZZY_SETS

Returns the type of the og fuzzy set computed before the time dependency.

membership(input: array, time: int = None) → array

Computes the membership of each sample and in each time for the fs.

Parameters:

input – array temporal_time x samples.

Time:

int. Time moment to compute the membership. If none, looks for a fixed time

Returns:

array temporal_time x samples.

type() → FUZZY_SETS

Returns the type of the fuzzy set. (temporal)

class ex_fuzzy.temporal.temporalFuzzyVariable(name: str, fuzzy_sets: list[temporalFS])

Bases: fuzzyVariable

Class to implement a temporal fuzzy variable.

compute_memberships(x: array, time: int = None) → dict

Computes the membership to each of the FS in the fuzzy variables.

Parameters:

• x – numeric value or array. Computes the membership to each of the IVFS in the fuzzy variables.

• time – int. Time moment to compute the membership.

Returns:

list of floats. Membership to each of the FS in the fuzzy variables.

fix_time(time: int) → None

Fixes the time of the temporal fuzzy variable.

Parameters:

time – int. Time moment to fix.

n_time_moments() → int

Returns the number of time moments of the temporal fuzzy variable.

Returns:

int. Number of time moments of the temporal fuzzy variable.

class ex_fuzzy.temporal.temporalMasterRuleBase(rule_base: list[MasterRuleBase], time_step_names: list[str] = None)

Bases: MasterRuleBase

This class is a temporal extension of the MasterRuleBase class. It includes a list of rule bases for each time step.

add_rule(rule: RuleSimple, consequent: int, time_step: int) → None

Adds a rule to the rule base of the given consequent.

Parameters:

• rule – rule to add.

• consequent – index of the rule base to add the rule.

• time_step – time step of the rule base to add the rule.

add_rule_base(rule_base: RuleBase, time: int) → None

Adds a rule base to the list of rule bases.

Parameters:

rule_base – rule base to add.

compute_firing_strenghts(X: array, time_moments: list[int], precomputed_truth=None) → array

Computes the firing strength of each rule for each sample.

Parameters:

X – array with the values of the inputs.

Returns:

array with the firing strength of each rule for each sample.

fuzzy_type() → FUZZY_SETS

Returns the correspoing type of the RuleBase using the enum type in the fuzzy_sets module.

Returns:

the corresponding fuzzy set type of the RuleBase.

get_consequents() → list[int]

Returns a list with the consequents of each rule base.

Returns:

list with the consequents of each rule base.

get_rulebase_matrix() → list[array]

Returns a list with the rulebases for each antecedent in matrix format.

Returns:

list with the rulebases for each antecedent in matrix format.

get_rulebases() → list[RuleBase]

Returns a list with all the rules.

Returns:

list

get_rules() → list[RuleSimple]

Returns a list with all the rules.

Returns:

list with all the rules.

get_scores() → array

Returns the dominance score for each rule in all the rulebases.

Returns:

array with the dominance score for each rule in all the rulebases.

print_rules(return_rules=False) → None

Print all the rules for all the consequents.

purge_rules(tolerance=0.001) → None

Delete the roles with a dominance score lower than the tolerance.

Parameters:

tolerance – tolerance to delete the rules.

winning_rule_predict(X: array, time_moments: int) → array

Returns the winning rule for each sample. Takes into account dominance scores if already computed.

Parameters:

X – array with the values of the inputs.

Returns:

array with the winning rule for each sample.