ClustererPipeline
Pipeline of transformers and a clusterer.
The ClustererPipeline compositor chains transformers and a single clusterer. The pipeline is constructed with a list of sktime transformers, plus a clusterer,
i.e., estimators following the BaseTransformer resp BaseClusterer interface.
- The transformer list can be unnamed - a simple list of transformers -
or string named - a list of pairs of string, estimator.
- For a list of transformers trafo1, trafo2, …, trafoN and a clusterer clst,
the pipeline behaves as follows:
- fit(X, y) - changes styte by running trafo1.fit_transform on X,
them trafo2.fit_transform on the output of trafo1.fit_transform, etc sequentially, with trafo[i] receiving the output of trafo[i-1], and then running clst.fit with X being the output of trafo[N], and y identical with the input to self.fit
- predict(X) - result is of executing trafo1.transform, trafo2.transform, etc
with trafo[i].transform input = output of trafo[i-1].transform, then running clst.predict on the output of trafoN.transform, and returning the output of clst.predict
- predict_proba(X) - result is of executing trafo1.transform, trafo2.transform,
etc, with trafo[i].transform input = output of trafo[i-1].transform, then running clst.predict_proba on the output of trafoN.transform, and returning the output of clst.predict_proba
- get_params, set_params uses sklearn compatible nesting interface
if list is unnamed, names are generated as names of classes if names are non-unique, f”_{str(i)}” is appended to each name string
where i is the total count of occurrence of a non-unique string inside the list of names leading up to it (inclusive)
- ClustererPipeline can also be created by using the magic multiplication
- on any clusterer, i.e., if my_clst inherits from BaseClusterer,
and my_trafo1, my_trafo2 inherit from BaseTransformer, then, for instance, my_trafo1 * my_trafo2 * my_clst will result in the same object as obtained from the constructor ClustererPipeline(clusterer=my_clst, transformers=[my_trafo1, my_trafo2])
- magic multiplication can also be used with (str, transformer) pairs,
as long as one element in the chain is a transformer
Quickstart
from sktime.clustering.compose import ClustererPipeline
estimator = ClustererPipeline(clusterer, transformers)Parameters(2)
- clusterersktime clusterer, i.e., estimator inheriting from BaseClusterer
- this is a “blueprint” clusterer, state does not change when fit is called
- transformerslist of sktime transformers, or
- list of tuples (str, transformer) of sktime transformers these are “blueprint” transformers, states do not change when fit is called
Examples
>>> from sktime.transformations.pca import PCATransformer
>>> from sktime.clustering.k_means import TimeSeriesKMeans
>>> from sktime.datasets import load_unit_test
>>> from sktime.clustering.compose import ClustererPipeline
>>> X_train, y_train = load_unit_test (split = "train")
>>> X_test, y_test = load_unit_test (split = "test")
>>> pipeline = ClustererPipeline (
... TimeSeriesKMeans (), [PCATransformer ()]
... )
>>> pipeline. fit (X_train, y_train) ClustererPipeline(
... )
>>> y_pred = pipeline. predict (X_test) Alternative construction via dunder method:
>>> pipeline = PCATransformer () * TimeSeriesKMeans ()