Transformer
HurstExponentTransformer
Transformer for calculating the Hurst exponent of a time series.
This transformer calculates the Hurst exponent, which is used to evaluate the auto-correlation properties of time series, particularly the degree of long-range dependence.
Quickstart
python
from sktime.transformations.hurst import HurstExponentTransformer
estimator = HurstExponentTransformer(lags: list [int ] | range | None=None, method: str='rs', min_lag: int=2, max_lag: int=100, fit_trend: str='c', confidence_level: float=0.95)Tags
Capabilities
- Unequal-length series
- Multivariate: Not supported
- Inverse transform: Not supported
- Missing values: Not supported
- Removes missing values: Not supported
- Equalizes series length: Not supported
Properties
- Input typescitype:transform-input
- Series
- Output typescitype:transform-output
- Primitives
- Label typescitype:transform-labels
- None
- Fit is emptyfit_is_empty
- No
- Keeps the time indextransform-returns-same-time-index
- No
- Requires Xrequires_X
- Yes
- Requires yrequires_y
- No
- X and y need the same indexX-y-must-have-same-index
- No
Parameters(6)
- lagsOptional[Union[List[int], range]], default=None
- The lags to use for calculation. If None, uses a range based on min_lag and max_lag.
- methodstr, default=’rs’
- The method to use for Hurst exponent calculation. Either ‘rs’ (rescaled range) or ‘dfa’ (detrended fluctuation analysis).
- min_lagint, default=2
- The minimum lag to use if lags is None.
- max_lagint, default=100
- The maximum lag to use if lags is None.
- fit_trendstr, default=’c’
- The trend component to include in the calculation.
- confidence_levelfloat, default=0.95
- The confidence level for the confidence interval calculation.
Examples
>>> from sktime.transformations.hurst import HurstExponentTransformer
>>> from sktime.datasets import load_airline
>>> y = load_airline ()
>>> transformer = HurstExponentTransformer ()
>>> y_transform = transformer. fit_transform (y)