API Reference

Provides functions to compute the LCSS similarity measure.

class lcsspy.lcss.LcssResult(lcss_measure: float, series_plot: Figure | None, sequence_plot: Figure | None)

Represents the outcome of the LCSS algorithm.

lcss_measure

LCSS similarity measure. Belongs to the range [0, 1].

Type:

float

series_plot

Figure object displaying the input time series and their matched elements (with green lines).

Type:

matplotlib.figure.Figure | None

sequence_plot

Figure object displaying the elements of the input time series that are part of the longest common subsequence.

Type:

matplotlib.figure.Figure | None

lcsspy.lcss.continuous_lcss(ts1: Series, ts2: Series, epsilon: float, delta: Timedelta, plot: bool) → LcssResult

Computes LCSS between time series with continuous time indexes.

Both time series can contain missing values in the form of numpy.nan. It’s expected that each series has sorted and unique timestamps.

Parameters:

• ts1 – First time series. The index property of ts1 must be of type DatetimeIndex. Must contain either integer or float values.

• ts2 – Second time series. The index property of ts2 must be of type DatetimeIndex. Must contain either integer or float values.

• epsilon – An upper bound to distances between time series values: elements can be matched only if their value distance is less than such threshold. Must be strictly positive.

• delta – An upper bound to distances between time series indexes: elements can be matched only if their index distance is less than or equal to such threshold. The value property of delta must be positive.

• plot – Indicates whether the function will return plots or not.

Returns:

A LcssResult object.

lcsspy.lcss.discrete_lcss(ts1: ndarray | Series, ts2: ndarray | Series, epsilon: float, delta: int, plot: bool) → LcssResult

Computes LCSS between time series with discrete time indexes.

Both time series can contain missing values in the form of numpy.nan. If a Series object is supplied, it will be converted to a ndarray, discarding its index in the process.

Parameters:

• ts1 – First time series. Must contain either integer or float values.

• ts2 – Second time series. Must contain either integer or float values.

• epsilon – An upper bound to distances between time series values: elements can be matched only if their value distance is less than such threshold. Must be strictly positive.

• delta – An upper bound to distances between time series indexes: elements can be matched only if their index distance is less than or equal to such threshold. Must be positive.

• plot – Indicates whether the function will return plots or not.

Returns:

A LcssResult object.