iSAX

class pyCFOFiSAX._isax.IndexableSymbolicAggregateApproximation(n_segments, alphabet_size_min=2, mean=0.0, std=1.0)

Indexable Symbolic Aggregate approXimation (iSAX) transformation.

First presented by J. Shieh & E. Keogh in iSAX: Indexing and Mining Terabyte Sized Time Series. Class that inherits the class PiecewiseAggregateApproximation proposed by Romain Tavenard in ``tslearn`` Available here <https://tslearn.readthedocs.io/en/stable/>`_.

Paramètres

  • n_segments (int) – The number of letters in the word sax

  • alphabet_size_min (int) – The minimum size of the Sax alphabet at initialization (2 default)

  • mean (float) – The average of the distribution of encoder sequences (0.0 default)

  • std (float) – The standard deviation of the distribution of encoder sequences (0.0 default)

_card_to_bkpt(max_cardinality)

Returns the breakpoints associated with the cardinations <= max_cardinality. The function calculates and stores the BKPT if they have never been calculated.

Paramètres

max_cardinality (int) – Maximum cardinality

Renvoie

Breakpoints associated with cardinality <= max_cardinality

Type renvoyé

dict

_card_to_bkpt_only(max_cardinality)

Returns the breakpoints associated with cardinality == max_cardinality. The function calculates and stores the BKPs if they have never been calculated.

Paramètres

max_cardinality (int) – cardinality

Renvoie

Breakpoints associated with cardinality == max_cardinality

Type renvoyé

list

_row_sax_word_array(ntss_tmp, bigger_cardinality, size_word)

Convert all sequences according to the different cardinality of the tree. For each cardinality, uses transform_sax().

Paramètres

  • ntss_tmp – The sequences to be analyzed

  • bigger_cardinality (int) – The greatest cardinality iSAX of the tree

  • size_word (int) – The size of the SAX sequences of the tree

Renvoie

SAX words from all ntss_tmp sequences according to all the cardinalities of the tree, a dict returning the cardinality index iSAX

Type renvoyé

numpy.ndarray, dict

_transform(X, card)

Transforms X data in parameter first into PAA and then in cardinate cardinality card.

Paramètres

  • X (numpy.ndarray) – Data to transform

  • card (int) – Cardinality to use for processing

Renvoie

Transformed data in SAX

Type renvoyé

numpy.ndarray

_transform_paa_to_isax(X_paa, card)

Transforms X_paa data into iSAX parameters according to cardinality card.

Paramètres

  • X_paa (numpy.ndarray) – PAA data to transform into iSAX

  • card (list) – Cardinalities to use for processing

Renvoie

Transformed data in SAX

Type renvoyé

numpy.ndarray

_transform_sax(X, card)

Transforms X data in parameter first into PAA and then in cardinate cardinality card.

Paramètres

  • X (numpy.ndarray) – Data to transform

  • card (int) – Cardinality to use for processing

Renvoie

Transformed data in SAX

Type renvoyé

numpy.ndarray

fit(X)

Prepares the data for encoding iSAX according to``PiecewiseAggregateApproximation``

Renvoie

Received data for encoding, defined by tslearn

Type renvoyé

numpy.ndarray of PiecewiseAggregateApproximation

fit_transform(X, card, **fit_params)

Prepares the X data provided in parameter for encoding``tslearn``. Then transforms the X data provided as a parameter first in PAA and then in cardinate cardinality card.

Paramètres

  • X (numpy.ndarray) – Data to transform

  • card (int) – Cardinality to use for processing

Renvoie

data transformed into SAX

Type renvoyé

numpy.ndarray

transform(X, card)

Prepares the X data provided in parameter for encoding tslearn. Then transforms X data in parameter first into PAA and then in cardinatlity of cardinality card.

Paramètres

  • X (numpy.ndarray) – Data to transform

  • card (int) – Cardinality to use for processing

Renvoie

Transformed data in SAX

Type renvoyé

numpy.ndarray

transform_paa(X)

Prepares the X data provided in parameter for encoding``tslearn``. Then transforms X data into parameter in PAA.

Paramètres

X (numpy.ndarray) – Data to transform

Renvoie

Transformed data in PAA

Type renvoyé

numpy.ndarray

transform_paa_to_isax(X_paa, card)

Prepares X_paa data provided as a parameter for encoding``tslearn``. Then transforms X_paa data into iSAX parameter according to cardinalities card.

Paramètres

  • X_paa (numpy.ndarray) – PAA data to transform into iSAX

  • card (list) – Cardinalities to use for processing

Renvoie

Transformed data in SAX

Type renvoyé

numpy.ndarray

transform_sax(X, card)

Prepares the X data provided in parameter for encoding``tslearn``. Then transforms X data in parameter first into PAA and then in cardinate cardinality``card``.

Paramètres

  • X (numpy.ndarray) – Data to transform

  • card (int) – Cardinality to use for processing

Renvoie

Transformed data in SAX

Type renvoyé

numpy.ndarray