Transformer
SummaryTransformer
Calculate summary value of a time series.
For univariate time series a combination of summary functions and quantiles of the input series are calculated. If the input is a multivariate time series then the summary functions and quantiles are calculated separately for each column.
Quickstart
python
from sktime.transformations.summarize import SummaryTransformer
estimator = SummaryTransformer(summary_function=('mean', 'std', 'min', 'max'), quantiles=(0.1, 0.25, 0.5, 0.75, 0.9), flatten_transform_index=True)Tags
Capabilities
- Multivariate
- Unequal-length series
- 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
- Yes
- 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(3)
- summary_functionstr, list, tuple, or None, default=(“mean”, “std”, “min”, “max”)
- If not None, a string, or list or tuple of strings indicating the pandas summary functions that are used to summarize each column of the dataset. Must be one of (“mean”, “min”, “max”, “median”, “sum”, “skew”, “kurt”, “var”, “std”, “mad”, “sem”, “nunique”, “count”). If None, no summaries are calculated, and quantiles must be non-None.
- quantilesstr, list, tuple or None, default=(0.1, 0.25, 0.5, 0.75, 0.9)
- Optional list of series quantiles to calculate. If None, no quantiles are calculated, and summary_function must be non-None.
- flatten_transform_indexbool, optional (default=True)
- if True, columns of return DataFrame are flat, by “variablename__feature” if False, columns are MultiIndex (variablename__feature) has no effect if return mtype is one without column names
Examples
>>> from sktime.transformations.summarize import SummaryTransformer
>>> from sktime.datasets import load_airline
>>> y = load_airline ()
>>> transformer = SummaryTransformer ()
>>> y_mean = transformer. fit_transform (y)