cartographer.mapper

Mapper algorithm implementation

class cartographer.mapper.Mapper(filterer=PCA(copy=True, iterated_power='auto', n_components=2, random_state=None, svd_solver='auto', tol=0.0, whiten=False), coverer=HyperRectangleCoverer(intervals=10, overlap=0.5), clusterer=DBSCAN(algorithm='auto', eps=0.5, leaf_size=30, metric='euclidean', min_samples=5, n_jobs=1, p=None), params=None)

Mapper algorithm implementation

Parameters:

filterer : scikit-learn transformer, default=PCA(n_components=2)

If y is not specified in the fit method the lense/filter values used will be the result of the application of this transform Pipeline to the input data will be considered.

coverer: instance of Coverer object, default=HyperRectangleCoverer()

clusterer: scikit-learn-like cluster estimator, default=DBSCAN()

params: dict of params for filterer, coverer and clusterer, default={}

Parameters to be passed to the filter, coverer and cluster objects

Methods

fit(X[, y])	Creates a Mapper graph for the input data
fit_predict(X[, y])	Performs clustering on X and returns cluster labels.
get_params([deep])	Get parameters for this estimator.
set_params(\*\*params)	Set the parameters of this estimator.

fit(X, y=None)

Creates a Mapper graph for the input data

If y is not None but a bidimensional array-like, its columns are used as lens/filter. If is None, the filters attribute is used.

Parameters:

X : array-like, shape=(n_samples, n_features)

Data on which the mapper algorithm will be applied.

y : array-like, shape=(n_samples, n_filters), optional

Lense/filter values for each samples, if not specified the class attribute filters will be used.