DOT：如何将集群放置在同一级别而不对这些集群中的节点进行同等排名？

【问题标题】：DOT: How to place clusters at same level without ranking equally the nodes in those clusters?DOT：如何将集群放置在同一级别而不对这些集群中的节点进行同等排名？
【发布时间】：2019-05-04 16:34:08
【问题描述】：

我有以下 DOT 代码。它描述了五个子图（集群）。我对两个较低集群的渲染以及集群的相对位置（三个在同一级别的顶部和两个在各自的底部的顺序）感到满意。但是，我希望上面三个集群中的节点根据它们节点的内部关系进行排名，类似于我在下面两个集群中得到的（如您所见，上面集群中的节点都呈现在一个单行，尽管其中一些之间存在隐藏边缘）。

我知道他们不是的原因是rank=same; 命令。但是，如果我删除它，我将失去集群的定位。我尝试了几个变体，使用命令的不同组合和其他命令组合，例如clusterrank=local;，但下面的结果保持最接近。 subgraph cluster ranking in dot 也没有得到我想要的。

digraph G {

    splines=line;
    size=1;
    ranksep=2;
    newrank=true;
    rankdir=BT

    subgraph cluster_z {
      label="Z";
      rank=same;

      "ZSF" [fillcolor = red];
      "ZTA" [fillcolor = red]; 
      "ZSS" [fillcolor = red];
      "ZIN" [fillcolor = red];
      "ZOW" [fillcolor = red];
      "ZNT" [fillcolor = red];
      "ZSS" [fillcolor = red];
      "ZCE" [fillcolor = red];
      "ZAY" [fillcolor = red];
      "ZNT" [fillcolor = red];
      "ZTA" [fillcolor = red];
      "ZTA" [fillcolor = red];
      "ZST" [fillcolor = red];
      "ZTO" [fillcolor = red];
      "ZON" [fillcolor = red];
      "ZPP" [fillcolor = red];
      "TNT" [fillcolor = red];
      "TCE" [fillcolor = red];
      "TNT" [fillcolor = red];


      "ZNT" -> "ZTA" [style=invis]
      "ZTA" -> "ZSF" [style=invis]
      "ZSF" -> "ZNT" [style=invis]

      "ZIN" -> "ZTA" [style=invis]
      "ZTA" -> "ZON" [style=invis]
      "ZON" -> "ZTA" [style=invis]
      "ZTA" -> "ZSF" [style=invis]
      "ZSF" -> "ZNT" [style=invis]
      "ZNT" -> "ZIN" [style=invis]

      "ZPP" -> "ZTA" [style=invis]
      "ZTA" -> "ZSF" [style=invis]
      "ZSF" -> "ZPP" [style=invis]

      "ZSF" -> "ZTA" [style=invis]
      "ZTA" -> "ZSF" [style=invis]

      "ZOW" -> "ZSS" [style=invis]
      "ZSS" -> "ZOW" [style=invis]

      "ZAY" -> "ZCE" [style=invis]
      "ZCE" -> "ZTA" [style=invis]
      "ZTA" -> "ZAY" [style=invis]

      "ZSF" -> "ZTA" [style=invis]
      "ZAY" -> "ZTA" [style=invis]
      "ZTA" -> "ZSF" [style=invis]

      "ZAY" -> "ZTA" [style=invis]
      "ZTA" -> "ZTA" [style=invis]
      "ZTA" -> "ZCE" [style=invis]
      "ZCE" -> "ZTA" [style=invis]
      "ZTA" -> "ZSF" [style=invis]
      "ZSF" -> "ZAY" [style=invis]

    }

    subgraph cluster_y {
      rank=same;
      label="Y";

      "YCY" [fillcolor = blue];
      "YES" [fillcolor = blue];
    }

    subgraph cluster_w {

      rank=same;
      label="W";

      "WER" [fillcolor = green];
      "WRT" [fillcolor = green];
    }

    subgraph cluster_o {

      label="O";
      "OOL" [fillcolor = white];
      "OOL" [fillcolor = white];
      "OIT" [fillcolor = white];
      "ONT" [fillcolor = white];
      "OGE" [fillcolor = white];
      "OTA" [fillcolor = white];
      "OTA" [fillcolor = white];
      "OTS" [fillcolor = white];
      "OTS" [fillcolor = white];
      "OCE" [fillcolor = white];
      "ORT" [fillcolor = white];
      "ORT" [fillcolor = white];
      "OON" [fillcolor = white];
      "OCT" [fillcolor = white];
      "OOL" [fillcolor = white];
      "OTO" [fillcolor = white];
      "OPE" [fillcolor = white];
      "OPY" [fillcolor = white];

      "OIT" -> "ORT" [style=invis]
      "ORT" -> "OON" [style=invis]
      "OON" -> "OPE" [style=invis]
      "OPE" -> "OIT" [style=invis]

      "OON" -> "OPE" [style=invis]
      "OPE" -> "OON" [style=invis]

      "OOL" -> "OOL" [style=invis]
      "OOL" -> "OOL" [style=invis]

      "OCE" -> "OON" [style=invis]
      "OON" -> "OCE" [style=invis]

      "OTA" -> "OON" [style=invis]
      "OON" -> "OTA" [style=invis]

      "OIT" -> "OTA" [style=invis]
      "OTA" -> "ORT" [style=invis]
      "ORT" -> "ORT" [style=invis]
      "ORT" -> "OON" [style=invis]
      "OON" -> "OIT" [style=invis]

      "OIT" -> "OTA" [style=invis]
      "OTA" -> "OTS" [style=invis]
      "OTS" -> "OON" [style=invis]
      "OON" -> "OPE" [style=invis]
      "OPE" -> "OIT" [style=invis]

    }

    subgraph cluster_e {
      label="E";

      "EUT" [fillcolor = grey];
      "EON" [fillcolor = grey];
      "ERT" [fillcolor = grey];
      "ERT" [fillcolor = grey];
      "EST" [fillcolor = grey];
      "EON" [fillcolor = grey];
      "EER" [fillcolor = grey];
      "ERE" [fillcolor = grey];
      "ETO" [fillcolor = grey];

    }

    "OIT" -> "ZSF";
    "ORT" -> "ZSF";
    "OON" -> "ZSF";
    "OPE" -> "ZSF";
    "EON" -> "ZSF";
    "EER" -> "ZSF";

    "OON" -> "ZTA";
    "OPE" -> "ZTA";
    "EER" -> "ZTA";

    "OOL" -> "ZSS";
    "OOL" -> "ZSS";

    "OGE" -> "ZIN";
    "EON" -> "ZIN";

    "OOL" -> "ZOW";

    "OTS" -> "ZNT";

    "OCT" -> "ZSS";

    "OCE" -> "ZCE";
    "OON" -> "ZCE";

    "OTA" -> "ZAY";
    "OON" -> "ZAY";

    "OIT" -> "ZNT";
    "EON" -> "ZNT";

    "OIT" -> "ZTA";
    "OTA" -> "ZTA";
    "ORT" -> "ZTA";
    "ORT" -> "ZTA";
    "OON" -> "ZTA";
    "EON" -> "ZTA";
    "ERE" -> "ZTA";

    "OIT" -> "ZTA";
    "OTA" -> "ZTA";
    "OTS" -> "ZTA";
    "OON" -> "ZTA";
    "OPE" -> "ZTA";
    "EON" -> "ZTA";
    "ERE" -> "ZTA";

    "OOL" -> "ZST";

    "OTO" -> "ZTO";

    "ONT" -> "ZON";
    "EON" -> "ZON";

    "OPY" -> "ZPP";
    "EER" -> "ZPP";


  { rank=same; "ZSF"; "YCY"; "WER" }
  { rank=same; "OPY"; "EER"}


}

【问题讨论】：

  • 您可能需要更详细地描述一下，最好是减少样本，您的意思是 我希望根据内部关系对上三个集群中的节点进行排名他们的节点，类似于我在较低的两个集群中得到的。您的代码非常复杂，需要花费大量时间才能理解，因此准确了解您想去哪里会很有帮助。

标签： graphviz dot


【解决方案1】：

我删除了所有重复的线条、不可见的边缘和颜色属性。

然后我给它一些颜色以显示我所做的事情。

来源是这样的：

digraph G {

newrank=true;
rankdir=BT;
ranksep=2;
splines=line;

subgraph cluster_z {
    label="Z";
    "TCE";
    "TNT";
    "ZAY";
    "ZCE";
    "ZIN";
    "ZNT";
    "ZON";
    "ZOW";
    "ZPP";
    "ZSF";
    "ZSS";
    "ZST";
    "ZTA";
    "ZTO";
}

subgraph cluster_y {
    rank=same;
    label="Y";
    "YCY";
    "YES";
}

subgraph cluster_w {
    rank=same;
    label="W";
    "WER";
    "WRT";
}

subgraph cluster_o {
    label="O";
    "OCE";
    "OCT";
    "OGE";
    "OIT";
    "ONT";
    "OOL";
    "OON";
    "OPE";
    "OPY";
    "ORT";
    "OTA";
    "OTO";
    "OTS";
}

subgraph cluster_e {
  label="E";
  "EUT";
  "ERT";
  "EST";
  "EON";
  "EER";
  "ERE";
  "ETO";
}

/* All the edges */
"EER" -> "ZPP";
"EER" -> "ZSF";
"EER" -> "ZTA";
"EON" -> "ZIN";
"EON" -> "ZNT";
"EON" -> "ZON";
"EON" -> "ZSF";
"EON" -> "ZTA";
"ERE" -> "ZTA";
"OCE" -> "ZCE";
"OCT" -> "ZSS";
"OGE" -> "ZIN";
"OIT" -> "ZNT";
"OIT" -> "ZSF";
"OIT" -> "ZTA";
"ONT" -> "ZON";
"OOL" -> "ZOW";
"OOL" -> "ZSS";
"OOL" -> "ZST";
"OON" -> "ZAY";
"OON" -> "ZCE";
"OON" -> "ZSF";
"OON" -> "ZTA";
"OPE" -> "ZSF";
"OPE" -> "ZTA";
"OPY" -> "ZPP";
"ORT" -> "ZSF";
"ORT" -> "ZTA";
"OTA" -> "ZAY";
"OTA" -> "ZTA";
"OTO" -> "ZTO";
"OTS" -> "ZNT";
"OTS" -> "ZTA";

/* 15 nodes from cluster_z and cluster_y and cluster_w */
{ rank = same;
    "TCE" [style = filled; color = green;]
    "TNT" [style = filled; color = green;]
    "WER" [style = filled; color = green;]
    "WRT" [style = filled; color = green;]
    "YCY" [style = filled; color = green;]
    "YES" [style = filled; color = green;]
    "ZAY" [style = filled; color = green;]
    "ZCE" [style = filled; color = green;]
    "ZIN" [style = filled; color = green;]
    "ZON" [style = filled; color = green;]
    "ZOW" [style = filled; color = green;]
    "ZPP" [style = filled; color = green;]
    "ZSS" [style = filled; color = green;]
    "ZST" [style = filled; color = green;]
    "ZTO" [style = filled; color = green;]
}

/* 10 nodes from cluster_o and cluster_e */
{ rank = same;
    "OCE" [style = filled; color = yellow;]
    "OCT" [style = filled; color = yellow;]
    "OGE" [style = filled; color = yellow;]
    "ONT" [style = filled; color = yellow;]
    "OOL" [style = filled; color = yellow;]
    "OPY" [style = filled; color = yellow;]
    "OTO" [style = filled; color = yellow;]
    "EON" [style = filled; color = yellow;]
    "EER" [style = filled; color = yellow;]
    "ERE" [style = filled; color = yellow;]
}

/* Gives cluster_z 2 rows of nodes. */
"ZSF" -> "ZTO" [color = red];

/* Gives cluster_o 4 rows of nodes. */
"OIT" -> "OON" -> "OPE" -> "ORT" [color = red];

}

虽然图像看起来像这样： 所以你可以随心所欲地摆弄它。

【讨论】：

  • 谢谢。您有什么想要改进的地方吗？
  • 我首先需要看看它在我的较大图表中的表现。我提出了一个不同的解决方案，将 5 个子图的 3/2 放置在每个子图中的两个级别上，并引入更多的不可见节点/边以使其放置正确。这些不可见节点有一定的原因，但您的解决方案完美地解决了所描述的问题。
  • 啊。我想知道所有这些不可见的节点和边缘是干什么用的。无论如何，我为这个示例所做的就是区分具有 3 个或更多箭头的节点和具有两个或更少箭头的节点。希望对您有所帮助。
猜你喜欢
  • 2011-10-13
  • 2014-11-02
  • 2019-09-19
  • 2014-06-13
  • 1970-01-01
  • 1970-01-01
  • 1970-01-01
  • 2017-01-06
  • 2020-07-08
相关资源
最近更新 更多
热门标签
Java Python linux javascript Mysql C# Docker 算法 前端 SpringBoot Redis Vue spring 设计模式 .net core .net kubernetes c++ 数据库 数据结构 大数据 js 机器学习 微服务 Android Go 程序员 面试 JVM ASP.net core 云原生 人工智能 后端 PHP git CSS golang k8s Nginx Django mybatis 深度学习 多线程 React 架构 devops 爬虫 云计算 Spring Boot LeetCode