Welcome to sciquence’s documentation!

Intro

Sciquence is a python module created especially to work with time series and other types of sequences. It mimics scikit-learn API, but introduces its own extensions as well.

http://www.timeseriesclassification.com/index.php

Installation

To install current bleeding-edge sciquence version, simply use command:

sudo pip install git+https://github.com/krzjoa/sciquence.git

sciquence API

sciquence.sequences

Cutting

seq(array) Cut input array into sequences consisting of the same elements
nseq(array) Returns sequences consisting of zeros
pseq(array) Returns sequences consisting of ones
specseq(array, element) Return sequences consisting of specific tag
seqi(array) Get list of sequences and corresponding list of indices
nseqi(array) Get list of negative sequences indices (consisting of zeroes)
pseqi(array) Get list of positive sequences indices (consisting of ones)
specseqi(array, elem) Get list of sequences indices, consisting of specific element
chunk(array, chunk_size) Split numpy array into chunks of equal length.

Comparing

lseq_equal(lseqa, lseqb) Compare two lists of ndarrays
shapes_equal(*arrays) Check if all the arrays have the same shape.
size_equal(*arrays, **kwargs) Check if all the arrays have the same length along the particular axis.

Sampling

random_slice(array_len, slice_length) Choose a random slice of given length

Sorting

parallel_sort(*arrays, **kwargs) Parallel sort.

Sliding window

wingen(X, window_size[, step, raw]) Generate subsequences from a single sequence.

Searching

mslc Given a length n real sequence, finds the consecutive subsequence of length at most U with the maximum sum in O(n) time.
longest_segment Find the longest subsequence which scores above a given threshold in O(n)
max_avg_seq Given a length n real sequence, finding the consecutive subsequence of length at least L with the maximum average can be done in O(n log L) time.

sciquence.similarities

Similarities

dtw(A, B, metric) Measure similarities between two sequences.
segmental_dtw Find similarities between two sequences.

sciquence.postprocessing

Binarization

ClasswiseBinarizer(thresholds) Performing binarization classwise.
binarize_classwise(X, thresholds) Binarization performed classwise.

sciquence.representation

Piecewise Aggregate Approximation

paa(sequence, window[, adjust]) Piecewise Aggregate Approximation

Symbolic Aggregate Approximation

sax(sequence, window[, alphabet_size, adjust]) Symbolic Aggregate Approximation.

sciquence.text_processing

Text to vector

Word2Idx() Class used for for transforming text data into word indices

sciquence.data_structures

Text to vector

MultiDict(**kwds) A dictionary, which allows to get multiple elements at once

Indices and tables