Question

我是R的新手，目前正致力于实施一种算法来修剪多图中的所有叶子，然后在结果图中完成每个循环，以便根据某些规则进一步简化它们。我目前正在测试几个数据集的算法，到目前为止它一直工作得很好，除了一个给我错误的特定数据集＆＃34;不能使用另一个图的边缘序列，＆＃34;我发现它很奇怪，因为它适用于更大的数据集。我花了最后几个小时试图检测错误无济于事。以下是我的代码：

pruneLeaf<-function(g){
   members<-membership(clusters(g,mode="strong"))
   strongly_connected=lapply(unique(members),function (x) 
                      induced.subgraph(g,which(members==x)))
   return (strongly_connected)
}

getNetflow<-function(g,node){
   #precondition: g is a graph and node is a vertex in graph g
   #postcondition: return the netflow from vertex node in graph g.
   inflow_list=incident_edges(g,node,mode="in")
   outflow_list=incident_edges(g,node,mode="out")
   inflow<-0
   outflow<-0
   for(item in edge_attr(g,"Quantity",inflow_list[[1]])){
       inflow<-inflow+item
       }
   for(item in edge_attr(g,"Quantity",outflow_list[[1]])){
       outflow<-outflow+item
      }
   return (inflow-outflow)
   }

modifyCycle<-function(g,node,netflow){
     #pre-condition:node of graph g has negative netflow
     #post-condition: eliminate transactions with the least priorities to           
     outflow_list=incident_edges(g,node,mode="out")
     if(length(outflow_list[[1]])==1){
        delete.edges(g,outflow_list[[1]])
       }
     else{
          priority_list=outflow_list[[1]][order(outflow_list[[1]]$Priority)]
          while(netflow<0){
          count=1
          edge=priority_list[count] #we begin with the lowest priority edge
          flow=edge$Quantity
          if(flow<= -netflow){ 
            netflow<-netflow+flow 
            g<-delete.edges(g,edge)
          }
          else{ #otherwise, we allow partial transfer to balance the netflow
              g<-set.edge.attribute(g,"Quantity",edge,flow+netflow)
              netflow<-0
              }
          count<-count+1
          }
          return (g)
          }
         }

 modifyGraph<-function(g){
      #precondition: g is an igraph object containing only cycles
      #postcondition: return an igraph with balanced flow at every cycle
      isDone=FALSE
      while(!isDone){
            vertices=V(g)
            isDone=TRUE
      for(node in vertices){
          netflow=getNetflow(g,node)
          if(netflow<0){
             g<-modifyCycle(g,node,netflow)
             isDone=FALSE
          }
         }
        }
       return (g)
      }

settlementGridlock<-function(g){
  #precondition: g is an igraph object
  #post-condition: return a list of graphs comprises of cyc 
  cyclic_graphs<-pruneLeaf(g)
  result<-lapply(cyclic_graphs,function(x) modifyGraph(x))
  return(result)
  }

对于相当长的帖子感到抱歉，但基本上有几个函数：

pruneLeaf：清除图表中的所有叶子并返回图表中强连接组件的列表
getNetflow：计算图表中给定节点的净流量
modifyCycle：根据给定的优先级删除边缘，以平衡
流入和流出给定节点
modifyGraph：此函数遍历图表以平衡流入和每个节点流出
settlementGridlock：此函数将modifyGraph应用于pruneLeaf生成的列表中的每个强连接组件（图）

错误似乎是由错误日志中给出的函数modifyCycle产生的：

Error in as.igraph.es(graph, index) : 
Cannot use an edge sequence from another graph. 
9 stop("Cannot use an edge sequence from another graph.") 
8 as.igraph.es(graph, index) 
7 i_set_edge_attr(graph = graph, name = name, index = index, value = value) 
6 set.edge.attribute(g, "Quantity", edge, flow + netflow) 
5 modifyCycle(g, node, netflow) 
4 modifyGraph(x) 
3 FUN(X[[i]], ...) 
2 lapply(cyclic_graphs, function(x) modifyGraph(x)) 
1 settlementGridlock(graph3)

无法使用其他图表中的边缘序列错误

0 个答案: