コード例 #1
0
ファイル: greedy_ter.py プロジェクト: Tetragen/ALGO1
def process_graph(graph_name):
    # load graph data
    with open(f"data/{graph_name}_neighbors", "rb") as f:
        neighbors = pickle.load(f)

    with open(f"data/{graph_name}_edges", "rb") as f:
        edges_list = pickle.load(f)

    # clean folder
    dir_name = f"images/{graph_name}"
    for filename in os.listdir(dir_name):
        if "initial" not in filename and "ter" in filename:
            path_to_file = os.path.join(dir_name, filename)
            os.remove(path_to_file)

    # print(neighbors)
    # size of the graph (number of nodes)
    nodes = neighbors.keys()
    n_nodes = len(nodes)
    """
        sort the nodes by degree
        aka the number of neighbors
        """
    sorted_nodes = sorted(neighbors,
                          key=lambda node: len(neighbors[node]),
                          reverse=True)
    """
        greedy algorithm
    """

    print('\n======')
    print('greedy algorithm')
    print('======')

    selected_nodes = list()
    dominated_nodes = list()
    step = 0

    def get_next_node(dominated_nodes, nodes):
        """
            function used to select a new node in the algorithm
        """
        """
            EDIT THIS FUNCTION
        """
        # built the reduced graph
        # consisting in the original graph deprived from dominated nodes
        reduced_graph = [node for node in nodes if node not in dominated_nodes]

        # build a new dictionary containing adjecency relations in that graph
        # we need to edit the dictionary of neighbors first.
        neighbors_reduced = neighbors.copy()
        """
            ADD LINES HERE
        """

        neighbors_in_reduced_graph = {
            node: neighbors_reduced[node]
            for node in reduced_graph
        }

        # sort that dictionary as before
        sorted_nodes_reduced = sorted(
            neighbors_in_reduced_graph,
            key=lambda node: len(neighbors_in_reduced_graph[node]),
            reverse=True)
        return sorted_nodes_reduced[0]

    while len(dominated_nodes) < n_nodes:
        selected_node = get_next_node(dominated_nodes, nodes)
        print(selected_nodes)

        # update graph
        selected_nodes.append(selected_node)
        print(f"\nadd {selected_node} to the dominating set")
        # update the list of dominated nodes
        dominated_nodes.append(selected_node)
        print(f"add {selected_node} to the list of dominated nodes")
        # print('neighbors : ')
        for neighbor in neighbors[selected_node]:
            if neighbor not in dominated_nodes:
                # update the list of not dominated nodes
                dominated_nodes.append(neighbor)
                print(f"add {neighbor} to the list of dominated nodes")
        # see how many more nodes we have to dominate
        print(f"still have to dominate {n_nodes-len(dominated_nodes)} nodes")
        step += 1
        plot_subset(step,
                    sorted_nodes,
                    edges_list,
                    dominated_nodes,
                    selected_nodes,
                    graph_name,
                    method="ter")
コード例 #2
0
ファイル: greedy_bis.py プロジェクト: Tetragen/ALGO1
def process_graph(graph_name):
    # load graph data
    with open(f"data/{graph_name}_neighbors", "rb") as f:
        neighbors = pickle.load(f)

    with open(f"data/{graph_name}_edges", "rb") as f:
        edges_list = pickle.load(f)

    # clean folder
    dir_name = f"images/{graph_name}"
    for filename in os.listdir(dir_name):
        if "initial" not in filename and "bis" in filename:
            path_to_file = os.path.join(dir_name, filename)
            os.remove(path_to_file)

    # print(neighbors)
    # size of the graph (number of nodes)
    nodes = neighbors.keys()
    n_nodes = len(neighbors)
    """
        sort the nodes by degree
        aka the number of neighbors
        """
    sorted_nodes = list(nodes)
    # sorted does not modify the original sequence
    # it returns a list
    # print(type(sorted_nodes))

    print('=====')
    print('sorted dictionary of neighbors by degree of the node')
    print('=====')

    for node in sorted_nodes:
        print(f"node  {node}")
        print(f"neighbors {neighbors[node]}")
    """
        greedy algorithm
    """

    print('\n======')
    print('greedy algorithm')
    print('======')

    selected_nodes = list()
    dominated_nodes = list()
    step = 0

    for node in sorted_nodes:
        step += 1
        # stop if the set is dominating
        if len(dominated_nodes) < n_nodes:
            # update our selected subset
            # selected_nodes.append(node)
            # print(f"\nadd {node} to the dominating set")
            # update the list of dominated nodes
            # if node not in dominated_nodes:
            #     dominated_nodes.append(node)
            # print(f"add {node} to the list of dominated nodes")
            # print('neighbors : ')
            # for neighbor in neighbors[node]:
            #     if neighbor not in dominated_nodes:
            # update the list of not dominated nodes
            #         dominated_nodes.append(neighbor)
            #         print(f"add {neighbor} to the list of dominated nodes")
            # see how many more nodes we have to dominate
            print(
                f"still have to dominate {n_nodes-len(dominated_nodes)} nodes")
            plot_subset(step,
                        sorted_nodes,
                        edges_list,
                        dominated_nodes,
                        selected_nodes,
                        graph_name,
                        method="bis")