probabilistic_model.learning.nyga_distribution ============================================== .. py:module:: probabilistic_model.learning.nyga_distribution Classes ------- .. autoapisummary:: probabilistic_model.learning.nyga_distribution.InductionStep probabilistic_model.learning.nyga_distribution.NygaDistribution Module Contents --------------- .. py:class:: InductionStep Class for performing induction in the NygaDistributions. .. py:attribute:: data :type: numpy.array The entire sorted and unique data points .. py:attribute:: cumulative_weights :type: numpy.array The cumulative log_weights of the samples in the dataset. .. py:attribute:: cumulative_log_weights :type: numpy.array The cumulative logarithmic log_weights of the samples in the dataset. .. py:attribute:: begin_index :type: int Included index of the first sample. .. py:attribute:: end_index :type: int Excluded index of the last sample. .. py:attribute:: nyga_distribution :type: NygaDistribution The Nyga Distribution to mount the quantile distributions into and read the parameters from. .. py:property:: variable The variable of the distribution. .. py:property:: min_samples_per_quantile The minimal number of samples per quantile. .. py:property:: min_likelihood_improvement The relative, minimal likelihood improvement needed to create a new quantile. .. py:method:: left_connecting_point() -> float Calculate the left connecting point. .. py:property:: number_of_samples The number of samples in the induction step. .. py:property:: total_weights The total sum of log_weights of the samples in the induction step. .. py:property:: total_log_weights The total sum of logarithmic log_weights of the samples in the induction step. .. py:method:: left_connecting_point_from_index(index) -> float Calculate the left connecting point given some beginning index. :param index: The index of the left datapoint. .. py:method:: right_connecting_point() -> float Calculate the right connecting point. .. py:method:: right_connecting_point_from_index(index) -> float Calculate the right connecting point given some ending index. :param index: The index of the right datapoint. .. py:method:: create_uniform_distribution() -> probabilistic_model.distributions.UniformDistribution Create a uniform distribution from this induction step. .. py:method:: create_uniform_distribution_from_indices(begin_index: int, end_index: int) -> probabilistic_model.distributions.UniformDistribution Create a uniform distribution from the datapoint at `begin_index` to the datapoint at `end_index`. :param begin_index: The index of the first datapoint. :param end_index: The index of the last datapoint. .. py:method:: sum_weights_from_indices(begin_index: int, end_index: int) -> float Sum the log_weights from `begin_index` to `end_index`. .. py:method:: sum_weights() Sum the log_weights of this induction step. .. py:method:: sum_log_weights_from_indices(begin_index: int, end_index: int) -> float Sum the logarithmic log_weights from `begin_index` to `end_index`. .. py:method:: sum_log_weights() Sum the logarithmic log_weights of this induction step. .. py:method:: compute_best_split() -> Tuple[float, Optional[int]] Compute the best split of the data. The best split of the data is computed by evaluating the log likelihood of every possible split and memorizing the best one. :return: The maximum log likelihood and the best split index. .. py:method:: log_likelihood_without_split() -> float Calculate the log likelihood without splitting. :return: The log likelihood without splitting. .. py:method:: log_likelihood_of_split_side(split_index: int, connecting_point: float) -> float Calculate the log likelihood of a split side. This method automatically determines if this is the left or right side of the split. :param split_index: The index of the split. :param connecting_point: The connecting point. :return: The log likelihood of the split. .. py:method:: construct_left_induction_step(split_index: int) -> typing_extensions.Self Construct the left induction step. :param split_index: The index of the split. .. py:method:: construct_right_induction_step(split_index: int) -> typing_extensions.Self Construct the right induction step. :param split_index: The index of the split. .. py:method:: improvement_is_good_enough(maximum_log_likelihood: float) -> bool Check if the improvement is good enough. :param maximum_log_likelihood: The improved maximum log likelihood. :return: Rather the improvement is good enough .. py:method:: induce() -> List[typing_extensions.Self] Perform one induction step. :return: The (possibly empty) list of new induction steps. .. py:class:: NygaDistribution Bases: :py:obj:`random_events.utils.SubclassJSONSerializer` A Nyga distribution is a way to learn a deterministic mixture of uniform distributions. .. py:attribute:: variable :type: random_events.variable.Continuous .. py:attribute:: min_likelihood_improvement :type: float :value: 0.01 The relative, minimal likelihood improvement needed to create a new quantile. .. py:attribute:: min_samples_per_quantile :type: int :value: 2 The minimal number of samples per quantile. .. py:attribute:: probabilistic_circuit :type: probabilistic_model.probabilistic_circuit.rx.probabilistic_circuit.ProbabilisticCircuit .. py:method:: fit(data: numpy.array, weights: Optional[numpy.array] = None) -> probabilistic_model.probabilistic_circuit.rx.probabilistic_circuit.ProbabilisticCircuit Fit the distribution to the data. :param data: The data to fit the distribution to. :param weights: The optional log_weights of the data points. :return: The fitted distribution. .. py:method:: to_json() -> Dict[str, Any] .. py:method:: _from_json(json_data: Dict[str, Any]) -> typing_extensions.Self :classmethod: Create a variable from a json dict. This method is called from the from_json method after the correct subclass is determined and should be overwritten by the respective subclass. :param data: The json dict :return: The deserialized object .. py:method:: empty_copy() -> typing_extensions.Self .. py:method:: from_uniform_mixture(mixture: probabilistic_model.probabilistic_circuit.rx.probabilistic_circuit.ProbabilisticCircuit) -> probabilistic_model.probabilistic_circuit.rx.probabilistic_circuit.ProbabilisticCircuit :staticmethod: Construct a Nyga Distribution from a mixture of uniform distributions. The mixture does not have to be deterministic. :param mixture: An arbitrary, univariate mixture of uniform distributions :return: A Nyga Distribution describing the same function. .. py:method:: all_union_of_mixture_points_with(other: typing_extensions.Self) Computes all possible union intervals of mixture points when combining two intervals. Returns: list of closed intervals representing all mixture points between distributions .. py:method:: event_of_higher_density(other: typing_extensions.Self, own_node_weights, other_node_weights) -> random_events.product_algebra.Event