Referencing to an Array inside of Object-CodePudding

There is following data Object:

{
  "station": {
    "id": {"name": "SNK"},
    "position": {"lat":0.00,"lon":0.00}
  },
  "datagram": [
    {
      "date": {"iso-8601": "1966-01-12"},
      "air": {
        "temp": {"mean": -17.8}
      },
      "precip": {"1d": 1}
    }
  ]
}

Data in question are inside of "datagram" array, and they are:

date.iso-8601,
air.temp.mean

So, is needed to indicate directly Array as data source.

I tried Transform with Calculate to create a new 'd' array:
(spec fragment)

data: {
  url: 'http://...',
  format: {type: 'json'}
},
transform: [
  {calculate: "datum.datagram", "as": "d"}
],
encoding: {
  x: {type: 'temporal', field: 'd.date.iso_date_time', title: 'Date'},
  y: {type: 'quantitative', field: 'd.air.temp.mean', title: '[°C]'}
}

It not work:

[email protected]:1 WARN Infinite extent for field "d.date.iso_date_time": [Infinity, -Infinity]

The Question is: How in general I should deal, in vanilla Vega-Lite, with extracting data table from an object?

Complete spec:

{
$schema: 'https://vega.github.io/schema/vega-lite/v5.json',
description: '',
width: 'container',
  data: {
      url:'',
      format: {type: 'json'}
  },
  transform: [
    {calculate: "datum.datagram", "as": "d"}
  ],
  mark: 'circle',
  encoding: {
     x: {type: 'temporal', field: 'd.time.iso_date_time', title: 'Date'},
     y: {type: 'quantitative', field: 'd.air.temp.mean', title: '[°C]'}
  }
};

CodePudding user response：

Here is an example.

{
  "$schema": "https://vega.github.io/schema/vega-lite/v5.json",
  "data": {
    "values": [
      {
        "station": {"id": {"name": "SNK"}, "position": {"lat": 0, "lon": 0}},
        "datagram": [
          {
            "date": {"iso-8601": "1966-01-12"},
            "air": {"temp": {"mean": -17.8}},
            "precip": {"1d": 1}
          }
        ]
      }
    ]
  },
  "transform": [
    {"calculate": "datum.datagram", "as": "d"},
    {"flatten": ["d"]},
    {"calculate": "datum.d.date['iso-8601']", "as": "x"},
      {"calculate": "datum.d.air.temp.mean", "as": "y"}
  ],
  "mark": "point",
  "encoding": {
    "x": {"field": "x", "type": "ordinal"},
    "y": {"field": "y", "type": "quantitative"}
  }
}