Labeller¶

class gtda.time_series.Labeller(width=10, func=<function std>, func_params=None, percentiles=None, n_steps_future=1)[source]¶

Target creation from sliding windows over a time series.

Useful to define a time series forecasting task in which labels are obtained from future values of the input time series, via the application of a function to time windows.

Parameters

width (int, optional, default: 10) – Width of each sliding window. Each window contains width + 1 objects from the original time series.
func (callable, optional, default: numpy.std) – Function to be applied to each window.
func_params (dict or None, optional, default: None) – Additional keyword arguments for func.
percentiles (list of real numbers between 0 and 100 inclusive, or None, optional, default: None) – If None, creates a target for a regression task. Otherwise, creates a target for an n-class classification task where n = len(percentiles) + 1.
n_steps_future (int, optional, default: 1) – Number of steps in the future for the predictive task.

thresholds\_

Values corresponding to each percentile, based on data seen in fit.

Type: list of floats or None if percentiles is None

Examples

>>> import numpy as np
>>> from gtda.time_series import Labeller
>>> # Create a time series
>>> X = np.arange(10).reshape(-1, 1)
>>> labeller = Labeller(width=2)
>>> # Fit and transform X
>>> X, y = labeller.fit_transform_resample(X, X)
>>> print(X)
[[1]
 [2]
 [3]
 [4]
 [5]
 [6]
 [7]
 [8]]
>>> print(y)
[0.81649658 0.81649658 0.81649658 0.81649658 0.81649658 0.81649658
 0.81649658 0.81649658]

__init__(width=10, func=<function std>, func_params=None, percentiles=None, n_steps_future=1)[source]¶: Initialize self. See help(type(self)) for accurate signature.

fit(X, y=None)[source]¶

Compute thresholds_ and return the estimator.

Parameters

X (ndarray of shape (n_samples,) or (n_samples, 1)) – Time series to build a target for.
y (None) – There is no need for a target in a transformer, yet the pipeline API requires this parameter.

Returns

self

Return type

object

fit_transform(X, y=None, **fit_params)¶

Fit to data, then transform it.

Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X.

Parameters

X (ndarray of shape (n_samples,) or (n_samples, 1)) – Time series to build a target for.
y (None) – There is no need for a target in a transformer, yet the pipeline API requires this parameter.

Returns

Xt – The cut input time series.

Return type

ndarray of shape (n_samples_new, 1)

fit_transform_resample(X, y, **fit_params)¶

Fit to data, then transform the input and resample the target. Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X ans a resampled version of y.

Parameters

X (ndarray of shape (n_samples, ..)) – Input data.
y (ndarray of shape (n_samples,)) – Target data.

Returns

Xt (ndarray of shape (n_samples, …)) – Transformed input.
yr (ndarray of shape (n_samples, …)) – Resampled target.

get_params(deep=True)¶

Get parameters for this estimator.

Parameters: deep (bool, default=True) – If True, will return the parameters for this estimator and contained subobjects that are estimators.
Returns: params – Parameter names mapped to their values.
Return type: mapping of string to any

resample(y, X=None)[source]¶

Resample y.

Parameters

y (ndarray of shape (n_samples,)) – Time series to build a target for.
X (None) – There is no need for X, yet the pipeline API requires this parameter.

Returns

yr – Target for the prediction task.

Return type

ndarray of shape (n_samples_new,)

set_params(**params)¶

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects (such as pipelines). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.

Parameters: **params (dict) – Estimator parameters.
Returns: self – Estimator instance.
Return type: object

transform(X, y=None)[source]¶

Cuts X so it is aligned with y.

Parameters

X (ndarray of shape (n_samples,) or (n_samples, 1)) – Time series to build a target for.
y (None) – There is no need for a target, yet the pipeline API requires this parameter.

Returns

Xt – The cut input time series.

Return type

ndarray of shape (n_samples_new, 1)

transform_resample(X, y)¶

Fit to data, then transform it.

Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X.

Parameters

X (ndarray of shape (n_samples, ..)) – Input data.
y (ndarray of shape (n_samples,)) – Target data.

Returns

Xt (ndarray of shape (n_samples, …)) – Transformed input.
yr (ndarray of shape (n_samples, …)) – Resampled target.