graph2vec.pyx 5.25 KiB
import hashlib
import json
import glob

import pandas as pd
import networkx as nx
from tqdm import tqdm
cimport numpy as np
import numpy.distutils.system_info as sysinfo

from joblib import Parallel, delayed
from gensim.models.doc2vec import Doc2Vec, TaggedDocument
from sklearn.metrics.pairwise import cosine_similarity

from ..base cimport Base
cimport cython


class WeisfeilerLehmanMachine:
    """
    Weisfeiler Lehman feature extractor class.
    """
    def __init__(self, graph, features, iterations):
        """
        Initialization method which executes feature extraction.
        
        Parameters
        ----------
        graph : nx.Graph
            graph
        features : dict
            Feature hash table.
        iterations : int
            number of WL iteration
        
        """

        self.iterations = iterations
        self.graph = graph
        self.features = features
        self.nodes = self.graph.nodes()
        self.extracted_features = [str(v) for k,v in features.items()]
        self.do_recursions()

    def do_a_recursion(self):
        """
        The method does a single WL recursion.
        
        Returns
        -------
        dict
            The hash table with extracted WL features.
        """

        new_features = {}
        for node in self.nodes:
            nebs = self.graph.neighbors(node)
            degs = [self.features[neb] for neb in nebs]
            features = "_".join([str(self.features[node])]+list(set(sorted([str(deg) for deg in degs]))))
            hash_object = hashlib.md5(features.encode())
            hashing = hash_object.hexdigest()
            new_features[node] = hashing
        self.extracted_features = self.extracted_features + list(new_features.values())
        return new_features

    def do_recursions(self):
        """
        The method does a series of WL recursions.
        """
        for iteration in range(self.iterations):
            self.features = self.do_a_recursion()
        

def dataset_reader(graph):
    """
    Function to extract features from a networkx graph
    
    Parameters
    ----------
    graph : nx.Graph
        graph
    
    Returns
    -------
    dict
        Features hash table.
    """

    features = dict(nx.degree(graph))

    features = {k:v for k,v, in features.items()}
    return graph, features


def feature_extractor(graph, ix, rounds):
    """
    Function to extract WL features from a graph
    
    Parameters
    ----------
    graph : nx.Graph
        graph
    ix : int
        index of the graph in the dataset
    rounds : int
        number of WL iterations
    
    Returns
    -------
    TaggedDocument 
        random walks
    """

    graph, features = dataset_reader(graph)
    machine = WeisfeilerLehmanMachine(graph,features,rounds)
    doc = TaggedDocument(words = machine.extracted_features , tags = ["g_{0}".format(ix)])
    return doc
        


def generate_model(graphs, iteration = 2, dimensions = 64, min_count = 5, down_sampling =  0.0001, learning_rate = 0.0001, epochs = 10, workers = 4 ):
    """
    Main function to read the graph list, extract features, learn the embedding and save it.
    
    Parameters
    ----------
    graphs : nx.Graph
        Input graph
    iteration : int, optional
        number of iteration (the default is 2)
    dimensions : int, optional
        output vector dimension (the default is 64)
    min_count : int, optional
        min count parameter of Doc2vec model (the default is 5)
    down_sampling : float, optional
        Down sampling rate for frequent features. (the default is 0.0001)
    learning_rate : float, optional
        Initial learning rate (the default is 0.0001, which [default_description])
    epochs : int, optional
        Number of epochs (the default is 10)
    workers : int, optional
        Number of workers (the default is 4)
    
    Returns
    -------
    [type]
        [description]
    """

    document_collections = Parallel(n_jobs = workers)(delayed(feature_extractor)(g, ix,iteration) for ix,g in tqdm(enumerate(graphs),desc="Extracting Features..."))
    graphs=[nx.relabel_nodes(g,{node:str(node) for node in list(g.nodes)},copy=True) for g in graphs]
    model = Doc2Vec(document_collections,
                    vector_size = dimensions,
                    window = 0,
                    min_count = min_count,
                    dm = 0,
                    sample = down_sampling,
                    workers = workers,
                    epochs = epochs,
                    alpha = learning_rate)
    return model

cdef class Graph2Vec(Base):
    """
    Based on :
    graph2vec: Learning distributed representations of graphs. 
    Narayanan, Annamalai and Chandramohan, Mahinthan and Venkatesan, Rajasekar and Chen, Lihui and Liu, Yang 
    MLG 2017, 13th International Workshop on Mining and Learning with Graphs (MLGWorkshop 2017)

    Orignal Code : https://github.com/benedekrozemberczki/graph2vec

    Modified by : Jacques Fize
    """

    def __init__(self):
        Base.__init__(self,0,True)

    @cython.boundscheck(False)
    cpdef np.ndarray compare(self,list listgs, list selected):
        # Selected is ignored
        model  = generate_model(listgs)
        vector_matrix = model.docvecs.vectors_docs
        cs = cosine_similarity(vector_matrix)
        return cs