javascript - D3.js 同时嵌套和汇总

Question

我希望将 csv 文件转换为特定的 JSON 格式以进行 D3.js 可视化。基本思想是在每个级别同时嵌套和汇总。换句话说，结构中每一层的 parent 都将包含其 child 的总和。

示例 csv 文件如下:

Country,State,City,Population,
"USA","California","Los Angeles",18500000,
"USA","California","San Diego",1356000,
"USA","California","San Francisco",837442,
"USA","Texas","Austin",885400,
"USA","Texas","Dallas",1258000,
"USA","Texas","Houston",2196000

我希望创建的 JSON 格式如下:

[
  {
    key: "USA",
    Population: 25032842,
    values: [
      {
        key: "California",
        Population: 20693442,
        values: [
          {
            key: "Los Angeles",
            Population: 18500000
          },
          {
            key: "San Diego",
            Population: 1356000
          },
          {
            key: "San Francisco",
            Population: 837442
          }
        ]
      },
      {
        key: "Texas",
        Population: 4339400,
        values: [
          {
            key: "Austin",
            Population: 885400
          },
          {
            key: "Dallas",
            Population: 1258000
          },
          {
            key: "Houston",
            Population: 2196000
          }
        ]
      }
    ]
  }
]

Answer 1

注意:这仅适用于 D3 v3。新版本 4 中情况略有变化，在访问汇总的返回值时需要进行一些调整。 "D3.js nesting and rollup at the same time in v4" 涵盖了这一点.

<小时/>

D3 没有内置功能来完成您正在寻找的功能。使用nest.rollup()将 trim 所有子节点，以用该函数的返回值替换它们。但是，编写一个小辅助函数可以通过两个步骤轻松完成:

1. 使用d3.nest()准备嵌套数据结构:

   var nested = d3.nest()
     .key(function(d) { return d.Country; })
     .key(function(d) { return d.State; })
     .rollup(function(cities) {
       return cities.map(function(c) {
         return {"City": c.City, "Population": +c.Population };
       });
     })
     .entries(data);
   

2. 循环遍历所有顶级节点以递归计算总和 所有 child 。

   // Recursively sum up children's values
   function sumChildren(node) {
     node.Population = node.values.reduce(function(r, v) {
       return r + (v.values ? sumChildren(v) : v.Population);
     },0);
     return node.Population;
   }
   
   // Loop through all top level nodes in nested data,
   // i.e. for all countries.
   nested.forEach(function(node) {
     sumChildren(node);
   });
   

这将为您提供所需的输出。请查看以下代码片段以了解其实际效果。

// Initialization
var csv = 'Country,State,City,Population\n' + 
'"USA","California","Los Angeles",18500000\n' + 
'"USA","California","San Diego",1356000\n' + 
'"USA","California","San Francisco",837442\n' + 
'"USA","Texas","Austin",885400\n' + 
'"USA","Texas","Dallas",1258000\n' + 
'"USA","Texas","Houston",2196000\n';

var data = d3.csv.parse(csv);

// Nesting the input using d3.nest()
var nested = d3.nest()
  .key(function(d) { return d.Country; })
  .key(function(d) { return d.State; })
  .rollup(function(cities) {
    return cities.map(function(c) {
      return {"City": c.City, "Population": +c.Population };
    });
  })
  .entries(data);

// Recursively sum up children's values
function sumChildren(node) {
  node.Population = node.values.reduce(function(r, v) {
    return r + (v.values ? sumChildren(v) : v.Population);
  },0);
  return node.Population;
}

// Loop through all top level nodes in nested data,
// i.e. for all countries.
nested.forEach(function(node) {
  sumChildren(node);
});

// Output. Nothing of interest below this line.
d3.select("body").append("div")
  .style("font-family", "monospace")
  .style("white-space", "pre")
  .text(JSON.stringify(nested,null,2));

<script src="https://cdnjs.cloudflare.com/ajax/libs/d3/3.4.11/d3.min.js"></script>