extendedMD.emd

extendedMD.emd.find_motifs_from_emd(ts, r, win_size, paa_size, alphabet_size, adaptive_break_points=True, z_threshold=0.01)

Returns the full list of motifs from either a multi-dimensional or a 1-dimensional time-series by running the extendedMD algorithm. If the time-series received is multi-dimensional (i.e. a pandas dataframe), then the algorithm starts by applying PCA to reduce it to a single dimension.

Parameters:

• ts (Union[1d array, DataFrame]) – original time-series
• r (float) – maximum distance to the center of the motif
• win_size (int) – size fo the sliding window that generated each sax word
• paa_size (int) – number of characters in a single sax word
• alphabet_size (int) – number of unique characters to use in the sax representation
• adaptive_break_points (bool) – Whether to use a representation with adaptive break-points
• z_threshold (float) – z_threshold for the znorm method from saxpy

Returns:

• motif_candidates_dic_list (list of dic) - list of motif dictionaries
• ts_1d (1d array) - 1-dimensional time-series either resulting from the PCA method or the original 1-dimensional time-series