opfython.math.general

General-based mathematical methods.

opfython.math.general.confusion_matrix(labels: Union[numpy.array, List[int]], preds: Union[numpy.array, List[int]])

Calculates the confusion matrix between true and predicted labels.

Parameters

  • labels – List or numpy array holding the true labels.

  • preds – List or numpy array holding the predicted labels.

Returns

The confusion matrix.

Return type

(np.array)

opfython.math.general.logger
opfython.math.general.normalize(array: numpy.array)

Normalizes an input array.

Parameters

array – Array to be normalized.

Returns

The normalized version (between 0 and 1) of the input array.

Return type

(np.array)

opfython.math.general.opf_accuracy(labels: Union[numpy.array, List[int]], preds: Union[numpy.array, List[int]])

Calculates the accuracy between true and predicted labels using OPF-style measure.

Parameters

  • labels – List or numpy array holding the true labels.

  • preds – List or numpy array holding the predicted labels.

Returns

The OPF accuracy measure between 0 and 1.

Return type

(float)

opfython.math.general.opf_accuracy_per_label(labels: Union[numpy.array, List[int]], preds: Union[numpy.array, List[int]])

Calculates the accuracy per label between true and predicted labels using OPF-style measure.

Parameters

  • labels – List or numpy array holding the true labels.

  • preds – List or numpy array holding the predicted labels.

Returns

The OPF accuracy measure per label between 0 and 1.

Return type

(float)

opfython.math.general.pre_compute_distance(data: numpy.array, output: str, distance: Optional[str] = 'log_squared_euclidean')

Pre-computes a matrix of distances based on an input data.

Parameters

  • data – Array of samples.

  • output – File to be saved.

  • distance – Distance metric to be used.

opfython.math.general.purity(labels: Union[numpy.array, List[int]], preds: Union[numpy.array, List[int]])

Calculates the purity measure of an unsupervised technique.

Parameters

  • labels – List or numpy array holding the true labels.

  • preds – List or numpy array holding the assigned labels by the clusters.

Returns

The purity measure.

Return type

(float)