Transformation¶

The documentation of the transformation module.

The pyts.transformation module includes transformation algorithms.

class pyts.transformation.BOSS(n_coefs, window_size, anova=False, norm_mean=True, norm_std=True, n_bins=4, quantiles=u'empirical', variance_selection=False, variance_threshold=0.0, numerosity_reduction=True)[source]¶

Bag-of-SFA Symbols.

Parameters:

n_coefs : None or int (default = None): The number of Fourier coefficients to keep. If n_coefs=None, all Fourier coefficients are returned. If n_coefs is an integer, the n_coefs most significant Fourier coefficients are returned if anova=True, otherwise the first n_coefs Fourier coefficients are returned. A even number is required (for real and imaginary values) if anova=False.
window_size : int: The size of the window.
anova : bool (default = False): If True, the Fourier coefficients selection is done via a one-way ANOVA test. If False, the first Fourier coefficients are selected.
norm_mean : bool (default = True): If True, center the data before scaling. If norm_mean=True and anova=False, the first Fourier coefficient will be dropped.
norm_std : bool (default = True): If True, scale the data to unit variance.
n_bins : int (default = 4): The number of bins. Ignored if quantiles='entropy'.
quantiles : {‘gaussian’, ‘empirical’} (default = ‘gaussian’): The way to compute quantiles. If ‘gaussian’, quantiles from a gaussian distribution N(0,1) are used. If ‘empirical’, empirical quantiles are used.
variance_selection : bool (default = False): If True, the Fourier coefficients with low variance are removed.
variance_threshold : float (default = 0.): Fourier coefficients with a training-set variance lower than this threshold will be removed. Ignored if variance_selection=False.
numerosity_reduction : bool (default = True): If True, numerosity reduction is applied: When the same word occurs several times in a row, only one instance of this word is kept.

Attributes:

vocabulary_ : dict: A mapping of features indices to terms.

Methods

`fit`(X[, y, overlapping])	Fit the model according to the given training data.
`fit_transform`(X[, y, overlapping])	Fit the data then transform it.
`get_params`([deep])	Get parameters for this estimator.
`set_params`(**params)	Set the parameters of this estimator.
`transform`(X)	Transform the provided data.

fit(X, y=None, overlapping=True)[source]¶

Fit the model according to the given training data.

Parameters:	X : array-like, shape = [n_samples, n_features] Training vector, where n_samples in the number of samples and n_features is the number of features. y : Ignored. overlapping : boolean (default = True) whether or not overlapping windows are used for the training phase.
Returns:	self : object

fit_transform(X, y=None, overlapping=True)[source]¶

Fit the data then transform it.

Parameters:	X : array-like, shape = [n_samples, n_features]
Returns:	X_new : sparse matrix, shape [n_samples, n_words] Document-term matrix.

transform(X)[source]¶

Transform the provided data.

Parameters:	X : array-like, shape = [n_samples, n_features]
Returns:	X_new : sparse matrix, shape [n_samples, n_words] Document-term matrix.

class pyts.transformation.WEASEL(n_coefs, window_sizes, norm_mean=True, norm_std=True, n_bins=4, variance_selection=False, variance_threshold=0.0, pvalue_threshold=0.9)[source]¶

Word ExtrAction for time SEries cLassification.