DENEB 3x3子图

发布于 2025-02-10 19:30:19 字数 1476 浏览 3 评论 0原文

我想在deneb中创建一个图表（我称其为3x3子图，但这可能不是正确的名称）：对于每个分类X和Y值，我想要一个项目名称的堆叠条形图。 [最终，将从AZ值添加颜色，基于组或层等强调。]我正在使用此示例数据

和下面的代码。目前，条是分层的，而不是堆叠。我认为我需要合并此转换（Measure CNT = Count（'table'[name]）），

  "transform": [
    {
      "stack": "cnt",
      "as": ["ymin", "ymax"],
      "groupby": ["x", "y"]
    }
  ]

但我不确定如何/何处将其插入下面的代码中。

{
  "data": {"name": "dataset"},
  "layer": [
    {
      "mark": {
        "type": "bar",
        "stroke": "black",
        "strokeWidth": 1,
        "tooltip": true
      }
    },
    {
      "mark": "text",
      "encoding": {
        "text": {
          "field": "Name",
          "type": "nominal"
        }
      }
    }
  ],
  "encoding": {
    "x": {
      "field": "x",
      "type": "nominal",
      "scale": {
        "domain": ["low", "med", "high"]
      },
      "axis": {
        "title": "X Level",
        "labelAngle": 0
      }
    },
    "y": {
      "field": "y",
      "type": "nominal",
      "scale": {
        "domain": ["high", "med", "low"]
      },
      "axis": {"title": "Y Level"}
    }
  }
}

需要明确的是，我的目标是删除名称的这种图表。

原文

I'm wanting to create a chart (I'm calling it a 3x3 subplot, but this may not be the correct name) in Deneb: for each categorical x and y value, I want a stacked bar chart of the project names. [Eventually, colors will be added from a z value, highlighting based on group or tier, etc.] I'm using this sample data

and the code below. Currently, the bars are layered rather than stacked. I think I need to incorporate this transformation (measure cnt = COUNT('Table'[name]))

  "transform": [
    {
      "stack": "cnt",
      "as": ["ymin", "ymax"],
      "groupby": ["x", "y"]
    }
  ]

but I'm not sure how/where to insert it into the code below.

{
  "data": {"name": "dataset"},
  "layer": [
    {
      "mark": {
        "type": "bar",
        "stroke": "black",
        "strokeWidth": 1,
        "tooltip": true
      }
    },
    {
      "mark": "text",
      "encoding": {
        "text": {
          "field": "Name",
          "type": "nominal"
        }
      }
    }
  ],
  "encoding": {
    "x": {
      "field": "x",
      "type": "nominal",
      "scale": {
        "domain": ["low", "med", "high"]
      },
      "axis": {
        "title": "X Level",
        "labelAngle": 0
      }
    },
    "y": {
      "field": "y",
      "type": "nominal",
      "scale": {
        "domain": ["high", "med", "low"]
      },
      "axis": {"title": "Y Level"}
    }
  }
}

To be clear, I'm aiming for this sort of chart where names have been removed.