pymfe.concept.MFEConcept

class pymfe.concept.MFEConcept[source]

Keep methods for metafeatures of Concept group.

The convention adopted for metafeature extraction related methods is to always start with ft_ prefix to allow automatic method detection. This prefix is predefined within _internal module.

All method signature follows the conventions and restrictions listed below:

  1. For independent attribute data, X means every type of attribute, N means Numeric attributes only and C stands for Categorical attributes only. It is important to note that the categorical attribute sets between X and C and the numerical attribute sets between X and N may differ due to data transformations, performed while fitting data into MFE model, enabled by, respectively, transform_num and transform_cat arguments from fit (MFE method).

  2. Only arguments in MFE _custom_args_ft attribute (set up inside fit method) are allowed to be required method arguments. All other arguments must be strictly optional (i.e., has a predefined default value).

  3. The initial assumption is that the user can change any optional argument, without any previous verification of argument value or its type, via kwargs argument of extract method of MFE class.

  4. The return value of all feature extraction methods should be a single value or a generic List (preferably a np.ndarray) type with numeric values.

There is another type of method adopted for automatic detection. It is adopted the prefix precompute_ for automatic detection of these methods. These methods run while fitting some data into an MFE model automatically, and their objective is to precompute some common value shared between more than one feature extraction method. This strategy is a trade-off between more system memory consumption and speeds up of feature extraction. Their return value must always be a dictionary whose keys are possible extra arguments for both feature extraction methods and other precomputation methods. Note that there is a share of precomputed values between all valid feature-extraction modules (e.g., class_freqs computed in module statistical can freely be used for any precomputation or feature extraction method of module landmarking).

__init__(*args, **kwargs)

Methods

__init__(*args, **kwargs)

ft_cohesiveness(N[, cohesiveness_alpha, ...])

Compute the improved version of the weighted distance, that captures how dense or sparse is the example distribution.

ft_conceptvar(N, y[, conceptvar_alpha, ...])

Compute the concept variation that estimates the variability of class labels among examples.

ft_impconceptvar(N, y[, ...])

Compute the improved concept variation that estimates the variability of class labels among examples.

ft_wg_dist(N[, wg_dist_alpha, ...])

Compute the weighted distance, that captures how dense or sparse is the example distribution.

precompute_concept_dist(N[, concept_dist_metric])

Precompute some useful things to support complexity measures.

classmethod ft_cohesiveness(N: ndarray, cohesiveness_alpha: float = 1.0, concept_dist_metric: str = 'euclidean', concept_distances: Optional[ndarray] = None) ndarray[source]

Compute the improved version of the weighted distance, that captures how dense or sparse is the example distribution.

Parameters
Nnp.ndarray

Numerical fitted data.

cohesiveness_alphafloat, optional

The alpha value to adjust the weight. The higher the alpha less is the effect of the weight in the computation.

concept_dist_metricstr, optional

Metric used to compute distance between each pair of examples. See cdist from scipy for more options. Used only if the argument concept_distances is None.

concept_distancesnp.ndarray, optional

Distance matrix of examples from N. Argument used to take advantage of precomputations.

Returns
np.ndarray

An array with the cohesiveness for each example.

References

1

Vilalta, R and Drissi, Y (2002). A Characterization of Difficult Problems in Classification. Proceedings of the 2002 International Conference on Machine Learning and Applications (pp. 133-138).

classmethod ft_conceptvar(N: ndarray, y: ndarray, conceptvar_alpha: float = 2.0, concept_dist_metric: str = 'euclidean', concept_minimum: float = 1e-09, concept_distances: Optional[ndarray] = None) ndarray[source]

Compute the concept variation that estimates the variability of class labels among examples.

Parameters
Nnp.ndarray

Numerical fitted data.

ynp.ndarray

Target attribute.

conceptvar_alphafloat, optional

The alpha value to adjust the weight. The higher the alpha less is the effect of the weight in the computation.

concept_dist_metricstr, optional

Metric used to compute distance between each pair of examples. See cdist from scipy for more options. Used only if the argument concept_distances is None.

concept_minimum: float, optional

This variable is the minimum value considered in the computation. It will be sum when necessary to avoid division by zero.

concept_distancesnp.ndarray, optional

Distance matrix of examples from N. Argument used to take advantage of precomputations.

Returns
np.ndarray

An array with the concept variation for each example.

References

1

Vilalta, R. (1999). Understanding accuracy performance through concept characterization and algorithm analysis. In Proceedings of the ICML-99 workshop on recent advances in meta-learning and future work (pp. 3-9).

classmethod ft_impconceptvar(N: ndarray, y: ndarray, impconceptvar_alpha: float = 1.0, concept_dist_metric: str = 'euclidean', concept_distances: Optional[ndarray] = None) ndarray[source]

Compute the improved concept variation that estimates the variability of class labels among examples.

Parameters
Nnp.ndarray

Numerical fitted data.

ynp.ndarray

Target attribute.

impconceptvar_alphafloat, optional

The alpha value to adjust the weight. The higher the alpha less is the effect of the weight in the computation.

concept_dist_metricstr, optional

Metric used to compute distance between each pair of examples. See cdist from scipy for more options. Used only if the argument concept_distances is None.

concept_distancesnp.ndarray, optional

Distance matrix of examples from N. Argument used to take advantage of precomputations.

Returns
np.ndarray

An array with the improved concept variation for each example.

References

1

Vilalta, R and Drissi, Y (2002). A Characterization of Difficult Problems in Classification. Proceedings of the 2002 International Conference on Machine Learning and Applications (pp. 133-138).

classmethod ft_wg_dist(N: ndarray, wg_dist_alpha: float = 2.0, concept_dist_metric: str = 'euclidean', concept_minimum: float = 1e-09, concept_distances: Optional[ndarray] = None) ndarray[source]

Compute the weighted distance, that captures how dense or sparse is the example distribution.

Parameters
Nnp.ndarray

Numerical fitted data.

wg_dist_alphafloat, optional

The alpha value to adjust the weight. The higher the alpha less is the effect of the weight in the computation.

concept_dist_metricstr, optional

Metric used to compute distance between each pair of examples. See cdist from scipy for more options. Used only if the argument concept_distances is None.

concept_minimumfloat, optional

This variable is the minimum value considered in the computation. It will be sum when necessary to avoid division by zero.

concept_distancesnp.ndarray, optional

Distance matrix of examples from N. Argument used to take advantage of precomputations.

Returns
np.ndarray

An array with the weighted distance for each example.

References

1

Vilalta, R. (1999). Understanding accuracy performance through concept characterization and algorithm analysis. In Proceedings of the ICML-99 workshop on recent advances in meta-learning and future work (pp. 3-9).

classmethod precompute_concept_dist(N: ndarray, concept_dist_metric: str = 'euclidean', **kwargs) Dict[str, Any][source]

Precompute some useful things to support complexity measures.

Parameters
Nnp.ndarray, optional

Numerical fitted data.

concept_dist_metricstr, optional

Metric used to compute distance between each pair of examples. See cdist from scipy for more options.

**kwargs

Additional arguments. May have previously precomputed before this method from other precomputed methods, so they can help speed up this precomputation.

Returns
dict
With following precomputed items:
  • concept_distances (np.ndarray): Distance matrix of examples from N.