0

我在 matlab 中有一个 MinCUT 算法。这是一个功能！我如何从我的邮件文件中调用它！我必须定义所有变量吗？还是我必须只定义输入？ 

function [MinCutGroupsList, MinCutWeight] = MinCut(SourceNodes, WeightedGraph)
%%% performs Min Cut algorithm described in "A Simple Min Cut Algorithm" by
%%% M. Stoer and F. Wagner.

%%% input - 
%%%     SourceNodes - a list of Nodes that are forced to be kept in one side of the cut.
%%%     WeightedGraph - symetric matrix of edge weights. Wi,j is the edge connecting Nodes i,j
%%%                     use Wi,j=0 or Wi,j == inf to indicate unconnected Nodes.
%%% output -
%%%   MinCutGroupsList - two lists of verices, SECOND one contains the sourve vertives
%%%   MinCutWeight - sum of weight of edges alosng the cut

    GraphDim = size(WeightedGraph,1);
    SourceNodes = SourceNodes(SourceNodes ~= 0); %remove zero Nodes

    %%% remove self edges and ZEROed ones
    WeightedGraph = WeightedGraph+diag(inf(1,GraphDim));
    % for ii = 1:GraphDim
    %     WeightedGraph(ii,ii) = inf;
    % end
    WeightedGraph(WeightedGraph == 0) = inf;

    %%%Merge all Source Vrtices to one, so they'll be unbreakable, descending order is VITAL!!!
    SourceNodes = sort(SourceNodes);
    GroupsList = zeros(GraphDim);   %each row are the Nodes melted into one vertex in the table.
    GroupsList(:,1) = 1:GraphDim;
    for ii=length(SourceNodes):-1:2;
        [WeightedGraph,GroupsList] = MeltTwoNodes(SourceNodes(1),SourceNodes(ii),WeightedGraph,GroupsList);
    end
    Split = GroupsList(:,1);

    %%% By now I have a weighted graph in which all seed Nodes are
    %%% merged into one vertex. Run Mincut algrithm on this graph
    [MinCutGroupsList_L, MinCutWeight] = MinCutNoSeed(WeightedGraph);

    %%% Convert Data so the seed Nodes will be reconsidered as different
    %%% Nodes and not one vertex.
    for ii = 1:2
        MinCutGroupsList(ii,:) = Local2GlobalIndices(MinCutGroupsList_L(ii,:), Split);
    end

    if (length(find(MinCutGroupsList(1,:) == SourceNodes(1))) == 1)
        SeedLocation = 1;
    else
        SeedLocation = 2;
    end
    MinCutGroupsList_withSeed = [MinCutGroupsList(SeedLocation,(MinCutGroupsList(SeedLocation,:)~=0)) SourceNodes(2:length(SourceNodes))];
    MinCutGroupsList_withSeed = sort(MinCutGroupsList_withSeed);
    MinCutGroupsList_withSeed = [MinCutGroupsList_withSeed zeros(1,GraphDim - length(MinCutGroupsList_withSeed))];

    MinCutGroupsList_NoSeed = MinCutGroupsList(3 - SeedLocation,(MinCutGroupsList(3 - SeedLocation,:)~=0));
    MinCutGroupsList_NoSeed = sort(MinCutGroupsList_NoSeed);
    MinCutGroupsList_NoSeed = [MinCutGroupsList_NoSeed zeros(1,GraphDim - length(MinCutGroupsList_NoSeed))];

    MinCutGroupsList = [MinCutGroupsList_NoSeed ; MinCutGroupsList_withSeed];

return
4

2 回答 2

1

将函数定义放在一个名为的文件中，该文件MinCut.m必须位于您当前的 matlab 路径中（例如，您当前的工作目录）。

然后你就可以调用函数编写了

 [MinCutGroupsList, MinCutWeight] = MinCut(yourSourceNodes, yourWeightedGraph);

还要考虑这个答案（Matlab：调用用户定义函数）。

于 2012-11-09T14:31:10.693 回答
0

你可以这样称呼它：

MinCut(SourceNodes, WeightedGraph)

或者如果您希望收集输出

[MinCutGroupsList, MinCutWeight] = MinCut(SourceNodes, WeightedGraph)
于 2012-11-09T14:31:27.323 回答