Source code for karateclub.graph_embedding.ldp

"""Local Degree Profile based embedding."""

import numpy as np
from karateclub.estimator import Estimator


[docs]class LDP(Estimator): r"""An implementation of `"LDP" <https://arxiv.org/abs/1811.03508>`_ from the ICLR Representation Learning on Graphs and Manifolds Workshop '19 paper "A Simple Yet Effective Baseline for Non-Attributed Graph Classification". The procedure calculates histograms of degree profiles. These concatenated histograms form the graph representations. Args: bins (int): Number of histogram bins. Default is 32. """ def __init__(self, bins: int = 32): self.bins = bins def _calculate_ldp(self, graph): """ Calculating the local degree profile features of a graph. Arg types: * **graph** *(NetworkX graph)* - A graph to be embedded. Return types: * **embedding** *(Numpy array)* - The embedding of a single graph. """ degrees = np.log( np.array([graph.degree[n] for n in range(graph.number_of_nodes())]) ) features = [] for n in range(graph.number_of_nodes()): nebs = [neb for neb in graph.neighbors(n)] degs = degrees[nebs] features.append([np.min(degs), np.max(degs), np.std(degs), np.mean(degs)]) features = np.concatenate([degrees.reshape(-1, 1), np.array(features)], axis=1) embedding = [] for i in range(features.shape[1]): x = features[:, i] emb = np.histogram(x, bins=self.bins, range=(0.0, 10.0))[0] embedding.append(emb) embedding = np.concatenate(embedding).reshape(-1) return embedding
[docs] def fit(self, graphs): """ Fitting an LDP model. Arg types: * **graphs** *(List of NetworkX graphs)* - The graphs to be embedded. """ graphs = self._check_graphs(graphs) self._embedding = [self._calculate_ldp(graph) for graph in graphs]
[docs] def get_embedding(self) -> np.array: r"""Getting the embedding of graphs. Return types: * **embedding** *(Numpy array)* - The embedding of graphs. """ return np.array(self._embedding)