How to set the domain from to the sum of a field to the sum of another field?

Question

I would like to calculate the sum of "SUM A" to set this as minimum scale value for all "X" axis. And the sum of "SUM B" to set als maximum scale for all "X" axis.

I can´t get my head around how to calculate the sum of the rows and use them in the "domain".

Hope the edit and example makes it more understandable.

I'm sorry if my first attempt to phrase the question wasn't clear.

{
  "data": { "values":[
{"EBENE_1": "A",    "EBENE_2":  "A",    "EBENE_3":  "A",    "EBENE_4":  "A" ,   "SUM A":    -70 ,   "SUM B":    41  },
{"EBENE_1": "A",    "EBENE_2":  "A",    "EBENE_3":  "B",    "EBENE_4":  "B" ,   "SUM A":    -22 ,   "SUM B":    81  },
{"EBENE_1": "A",    "EBENE_2":  "B",    "EBENE_3":  "C",    "EBENE_4":  "C" ,   "SUM A":    -31 ,   "SUM B":    89  },
{"EBENE_1": "A",    "EBENE_2":  "B",    "EBENE_3":  "D",    "EBENE_4":  "D" ,   "SUM A":    -28 ,   "SUM B":    8   },
{"EBENE_1": "A",    "EBENE_2":  "C",    "EBENE_3":  "E",    "EBENE_4":  "E" ,   "SUM A":    -39 ,   "SUM B":    55  },
{"EBENE_1": "A",    "EBENE_2":  "C",    "EBENE_3":  "E",    "EBENE_4":  "F" ,   "SUM A":    -78 ,   "SUM B":    78  },
{"EBENE_1": "A",    "EBENE_2":  "C",    "EBENE_3":  "F",    "EBENE_4":  "G" ,   "SUM A":    -99 ,   "SUM B":    88  },
{"EBENE_1": "A",    "EBENE_2":  "C",    "EBENE_3":  "F",    "EBENE_4":  "H" ,   "SUM A":    -85 ,   "SUM B":    67  }
]
},
"facet": {
    "row": {
      "field": "EBENE_1",
      "header": {"labelAngle": 0, "title": null, "labelAnchor": "start", "labelLimit": 50},
      "type": "nominal"
    }
  },
  "spec": {
    "hconcat": [
      {
        "height": 10,
        "mark": {"type": "bar"},
   "encoding": {
    "x": {"field": "SUM A", "aggregate": "sum"},
    "y": {"field": "EBENE_1", "title": null}}
      },
      {
        "facet": {
          "row": {
            "field": "EBENE_2",
      "header": {"labelAngle": 0, "title": null, "labelAnchor": "start", "labelLimit": 50},
            "type": "nominal"
          }
        },
        "spec": {
          "hconcat": [
            {
              "height": 10,
              "mark": {"type": "bar"},
   "encoding": {
    "x": {"field": "SUM A", "aggregate": "sum"},
    "y": {"field": "EBENE_2", "title": null}}
            },
            {
              "facet": {
                "row": {
                  "field": "EBENE_3",
      "header": {"labelAngle": 0, "title": null, "labelAnchor": "start", "labelLimit": 50},
                  "type": "nominal"
                }
              },
              "spec": {
                "hconcat": [
                  {
                    "height": 10,
                    "mark": {
                      "type": "bar"
                    },
   "encoding": {
    "x": {"field": "SUM A", "aggregate": "sum"},
    "y": {"field": "EBENE_3", "title": null}}
                  },
                  {
                    "facet": {
                      "row": {
                        "field": "EBENE_4",
      "header": {"labelAngle": 0, "title": null, "labelAnchor": "start", "labelLimit": 50},
                        "type": "nominal"
                      }
                    },
                    "spec": {
                      "hconcat": [
                        {
                          "height": 10,
                          "mark": {
                            "type": "bar"
                          },
   "encoding": {
    "x": {"field": "SUM A", "aggregate": "sum"},
    "y": {"field": "EBENE_4", "title": null}}
                        }
                      ]
                    }
                  }
                ]
              }
            }
          ]
        }
      }
    ]
  }
}

Please share a minimum reproducible example of the problem you're facing and upir expected results. — Davide Bacci, Mar 03 '23 at 16:56
ChatGPT can't help if you can't even articulate your question. Same thing applies to the voluntary helpers at StackOverflow. — Peter, Mar 03 '23 at 22:20
Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. — Fernando Barbosa, Mar 05 '23 at 02:18

score 1 · Answer 1 · answered Mar 10 '23 at 15:59

I struggle with these too. Here's my view of the problem and my workaround.

Scale domains want hard-coded values. To my knowledge, the only way to make a scale domain dynamic is to use signals. Vega-lite doesn't have many ways to go from datum to signal. Params are the only way I know. They are funky to use.

Step 1: Get your scale max and min in a column
We need a max and min. If you have it in your data, that will work. Otherwise, you need to calculate them with a transform. I use a joinaggregate transform so that it adds a column to the data. In your data, you are using sum. In most cases, this will be "op": "max" and "op": "min".

  "transform": [
    {"joinaggregate": [{"op": "sum", "field": "SUM_A", "as": "SumA"}]},
    {"joinaggregate": [{"op": "sum", "field": "SUM_B", "as": "SumB"}]}
  ],

Step 2: Convert datum to signal
To get the data values into signals, use params. The hacky notation here uses data_0 (which should be the dataset after the transforms), the first row [0], and the columns we just added [SumA] and [SumB].

  "params": [
    {
      "name": "lowerLimit",
      "expr": "data('data_0')[0]['SumA']"
    },
    {
      "name": "upperLimit",
      "expr": "data('data_0')[0]['SumB']"
    }

  ],

Check the signal viewer to make sure these values are as expected.

Step 3: Add signal values to scale domain
Reference the signal values in an expression object {"expr": "lowerLimit"}. Note that you just use the signal name.

          "x": {
            "field": "SUM_A", 
            "aggregate": "sum", 
            "scale": {
              "domain": [{"expr": "lowerLimit"},{"expr": "upperLimit"}]
            }
          },

Result:

Here's the full spec: Open the Chart in the Vega Editor

{
  "data": {
    "values": [
      {
        "EBENE_1": "A",
        "EBENE_2": "A",
        "EBENE_3": "A",
        "EBENE_4": "A",
        "SUM_A": -70,
        "SUM_B": 41
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "A",
        "EBENE_3": "B",
        "EBENE_4": "B",
        "SUM_A": -22,
        "SUM_B": 81
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "B",
        "EBENE_3": "C",
        "EBENE_4": "C",
        "SUM_A": -31,
        "SUM_B": 89
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "B",
        "EBENE_3": "D",
        "EBENE_4": "D",
        "SUM_A": -28,
        "SUM_B": 8
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "C",
        "EBENE_3": "E",
        "EBENE_4": "E",
        "SUM_A": -39,
        "SUM_B": 55
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "C",
        "EBENE_3": "E",
        "EBENE_4": "F",
        "SUM_A": -78,
        "SUM_B": 78
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "C",
        "EBENE_3": "F",
        "EBENE_4": "G",
        "SUM_A": -99,
        "SUM_B": 88
      },
      {
        "EBENE_1": "A",
        "EBENE_2": "C",
        "EBENE_3": "F",
        "EBENE_4": "H",
        "SUM_A": -85,
        "SUM_B": 67
      }
    ]
  },
  "transform": [
    {"joinaggregate": [{"op": "sum", "field": "SUM_A", "as": "SumA"}]},
    {"joinaggregate": [{"op": "sum", "field": "SUM_B", "as": "SumB"}]}
  ],
  "params": [
    {
      "name": "lowerLimit",
      "expr": "data('data_0')[0]['SumA']"
    },
    {
      "name": "upperLimit",
      "expr": "data('data_0')[0]['SumB']"
    }

  ],
  "facet": {
    "row": {
      "field": "EBENE_1",
      "header": {
        "labelAngle": 0,
        "title": null,
        "labelAnchor": "start",
        "labelLimit": 50
      },
      "type": "nominal"
    }
  },
  "spec": {
    "hconcat": [
      {
        "height": 10,
        "mark": {"type": "bar"},
        "encoding": {
          "x": {
            "field": "SUM_A", 
            "aggregate": "sum", 
            "scale": {
              "domain": [{"expr": "lowerLimit"},{"expr": "upperLimit"}]
            }
          },
          "y": {"field": "EBENE_1", "title": null}
        }
      },
      {
        "facet": {
          "row": {
            "field": "EBENE_2",
            "header": {
              "labelAngle": 0,
              "title": null,
              "labelAnchor": "start",
              "labelLimit": 50
            },
            "type": "nominal"
          }
        },
        "spec": {
          "hconcat": [
            {
              "height": 10,
              "mark": {"type": "bar"},
              "encoding": {
                "x": {"field": "SUM_A", "aggregate": "sum", 
            "scale": {
              "domain": [{"expr": "lowerLimit"},{"expr": "upperLimit"}]
            }},
                "y": {"field": "EBENE_2", "title": null}
              }
            },
            {
              "facet": {
                "row": {
                  "field": "EBENE_3",
                  "header": {
                    "labelAngle": 0,
                    "title": null,
                    "labelAnchor": "start",
                    "labelLimit": 50
                  },
                  "type": "nominal"
                }
              },
              "spec": {
                "hconcat": [
                  {
                    "height": 10,
                    "mark": {"type": "bar"},
                    "encoding": {
                      "x": {"field": "SUM_A", "aggregate": "sum", 
            "scale": {
              "domain": [{"expr": "lowerLimit"},{"expr": "upperLimit"}]
            }},
                      "y": {"field": "EBENE_3", "title": null}
                    }
                  },
                  {
                    "facet": {
                      "row": {
                        "field": "EBENE_4",
                        "header": {
                          "labelAngle": 0,
                          "title": null,
                          "labelAnchor": "start",
                          "labelLimit": 50
                        },
                        "type": "nominal"
                      }
                    },
                    "spec": {
                      "hconcat": [
                        {
                          "height": 10,
                          "mark": {"type": "bar"},
                          "encoding": {
                            "x": {"field": "SUM_A", "aggregate": "sum", 
            "scale": {
              "domain": [{"expr": "lowerLimit"},{"expr": "upperLimit"}]
            }},
                            "y": {"field": "EBENE_4", "title": null}
                          }
                        }
                      ]
                    }
                  }
                ]
              }
            }
          ]
        }
      }
    ]
  }
}

I'll test this tomorrow, thank you for this in depth explanation. — Fabian Storck, Mar 12 '23 at 10:26
The thing that was missing for me was "zero": "true" in the domain. Thank you very much for the help! Saw this question a couple of times here, but without a solution. — Fabian Storck, Mar 14 '23 at 14:02
I'm hoping to figure out an easier way, but in the meantime this works. Maybe a future version of Vega will make this easier. It's a very common challenge, especially for Deneb/Power BI users. — TheRizza, Mar 14 '23 at 20:41

score 0 · Accepted Answer · answered Mar 19 '23 at 12:58

Another solution to my problem was suggested by the creator of Deneb, as there are at the moment problems using data expressions and repeat/concate/facet:

{
  "data": {"name": "dataset"},
    "resolve": {
      "scale": {
        "x": "shared"
      }
  },
  "repeat": {
    "layer": ["STORNO", "EINGANG"],
    "column": [
      "EBENE_1",
      "EBENE_2",
      "EBENE_3",
      "EBENE_4"
    ]
  },
  "spec": {
    "height": 400,
    "width": 250,
    "mark": {"type": "bar"},
    "encoding": {
      "x": {
        "field": {"repeat": "layer"},
        "aggregate": "sum"
      },
      "y": {
        "field": {"repeat": "column"}
      }
    }
  }
}

How to set the domain from to the sum of a field to the sum of another field?

2 Answers2