giotto.diagrams.PairwiseDistance

class giotto.diagrams.PairwiseDistance(metric='landscape', metric_params=None, order=2.0, n_jobs=None)

Distances between pairs of persistence diagrams, constructed from the distances between their respective subdiagrams with constant homology dimension.

Given two collections of persistence diagrams consisting of birth-death-dimension triples [b, d, q], a collection of distance matrices or a single distance matrix between pairs of diagrams is calculated according to the following steps:

1. All diagrams are partitioned into subdiagrams corresponding to distinct homology dimensions.

2. Pairwise distances between subdiagrams of equal homology dimension are calculated according to the parameters metric and metric_params. This gives a collection of distance matrices, \(\mathbf{D} = (D_{q_1}, \ldots, D_{q_n})\).

3. The final result is either \(\mathbf{D}\) itself as a three-dimensional array, or a single distance matrix constructed by taking norms of the vectors of distances between diagram pairs.

Parameters

metric'bottleneck' | 'wasserstein' | 'landscape' | 'betti' | 'heat', optional, default: 'landscape'

Distance or dissimilarity function between subdiagrams:

• 'bottleneck' and 'wasserstein' refer to the identically named perfect-matching–based notions of distance.

• 'landscape' refers to the \(L^p\) distance between persistence landscapes.

• 'betti' refers to the \(L^p\) distance between Betti curves.

• 'heat' refers to the \(L^p\) distance between Gaussian-smoothed diagrams.

metric_paramsdict or None, optional, default: None

Additional keyword arguments for the metric function:

• If metric == 'bottleneck' the only argument is delta (float, default: 0.01). When equal to 0., an exact algorithm is used; otherwise, a faster approximate algorithm is used.

• If metric == 'wasserstein' the available arguments are p (int, default: 2) and delta (float, default: 0.01). Unlike the case of 'bottleneck', delta cannot be set to 0. and an exact algorithm is not available.

• If metric == 'betti' the available arguments are p (float, default: 2.) and n_values (int, default: 100).

• If metric == 'landscape' the available arguments are p (float, default: 2.), n_values (int, default: 100) and n_layers (int, default: 1).

• If metric == 'heat' the available arguments are p (float, default: 2.), sigma (float, default: 1.) and n_values (int, default: 100).

orderfloat or None, optional, default: 2.

If None, transform returns for each pair of diagrams a vector of distances corresponding to the dimensions in homology_dimensions_. Otherwise, the \(p\)-norm of these vectors with \(p\) equal to order is taken.

n_jobsint or None, optional, default: None

The number of jobs to use for the computation. None means 1 unless in a joblib.parallel_backend context. -1 means using all processors.

Attributes

effective_metric_params_dict

Dictionary containing all information present in metric_params as well as on any relevant quantities computed in fit.

homology_dimensions_list

Homology dimensions seen in fit, sorted in ascending order.

See also

Amplitude, BettiCurve, PersistenceLandscape, HeatKernel, giotto.homology.VietorisRipsPersistence

Notes

To compute distances without first splitting the computation between different homology dimensions, data should be first transformed by an instance of ForgetDimension.

Hera is used as a C++ backend for computing bottleneck and Wasserstein distances between persistence diagrams. Python bindings were modified for performance from the Dyonisus 2 package.

Methods

fit(self, X[, y])	Store all observed homology dimensions in homology_dimensions_ and compute effective_metric_params.
fit_transform(self, X[, y])	Fit to data, then transform it.
get_params(self[, deep])	Get parameters for this estimator.
set_params(self, \*\*params)	Set the parameters of this estimator.
transform(self, X[, y])	Computes a distance or vector of distances between the diagrams in X and the diagrams seen in fit.

__init__(self, metric='landscape', metric_params=None, order=2.0, n_jobs=None)

Initialize self. See help(type(self)) for accurate signature.

fit(self, X, y=None)

Store all observed homology dimensions in homology_dimensions_ and compute effective_metric_params. Then, return the estimator.

This method is there to implement the usual scikit-learn API and hence work in pipelines.

Parameters

Xndarray, shape (n_samples_fit, n_features, 3)

Input data. Array of persistence diagrams, each a collection of triples [b, d, q] representing persistent topological features through their birth (b), death (d) and homology dimension (q).

yNone

There is no need for a target in a transformer, yet the pipeline API requires this parameter.

Returns

selfobject

fit_transform(self, 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

Xnumpy array of shape [n_samples, n_features]

Training set.

ynumpy array of shape [n_samples]

Target values.

Returns

X_newnumpy array of shape [n_samples, n_features_new]

Transformed array.

get_params(self, deep=True)

Get parameters for this estimator.

Parameters

deepboolean, optional

If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns

paramsmapping of string to any

Parameter names mapped to their values.

set_params(self, **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.

Returns

self

transform(self, X, y=None)

Computes a distance or vector of distances between the diagrams in X and the diagrams seen in fit.

Parameters

Xndarray, shape (n_samples, n_features, 3)

Input data. Array of persistence diagrams, each a collection of triples [b, d, q] representing persistent topological features through their birth (b), death (d) and homology dimension (q).

yNone

There is no need for a target in a transformer, yet the pipeline API requires this parameter.

Returns

Xtndarray, shape (n_samples_fit, n_samples, n_homology_dimensions) if order is None, else (n_samples_fit, n_samples)

Distance matrix or collection of distance matrices between diagrams in X and diagrams seen in fit. In the second case, index i along axis 2 corresponds to the i-th homology dimension in homology_dimensions_.