Interval scale produces wrong ticks and labels

[spren9er]

According to Observable Plot docs, interval scale can take numbers:

The option can also be specified as a number, in which case it will be promoted to a numeric interval with the given step.

In the following example, I apply an interval transform and an interval scale (both given by a number) to pre-binned data. Resulting plot has wrong x-axis ticks and labels.

Example

See also Observable Notebook.

Given data (already binned with bin width 0.05 or bin step -20)

data = [
  { "x": 0.15, "y": 1 },
  { "x": 0.25, "y": 5 },
  { "x": 0.35, "y": 85 },
  { "x": 0.4,  "y": 196 },
  { "x": 0.45, "y": 268 },
  { "x": 0.5,  "y": 380 },
  { "x": 0.55, "y": 354 },
  { "x": 0.6,  "y": 658 },
  { "x": 0.65, "y": 7382 },
  { "x": 0.7,  "y": 45588 },
  { "x": 0.75, "y": 67028 },
  { "x": 0.8,  "y": 52965 },
  { "x": 0.9,  "y": 95069 },
  { "x": 0.95, "y": 30982 }
]

With Interval transform, but without Interval scale

Plot.plot({
  marks: [
    Plot.rectY(data, { x: "x", y: "y", interval: 0.05 })
  ]
})

This looks good. Now, I want to add an interval scale:

With Interval transform, with Interval scale

Plot.plot({
  x: {
    interval: 0.05
  },
  marks: [
    Plot.rectY(data, { x: "x", y: "y", interval: 0.05 })
  ]
})

This plot has a wrong x-axis:

• Ticks are off by 0.05
• 0.3 and 0.8 are each shown twice
• 0.2 and 0.4 are missing

Increasing the number of ticks produces a correct plot:

With Interval transform, with Interval scale, and with explicit ticks

Plot.plot({
  x: {
    interval: 0.05,
    ticks: 18
  },
  marks: [
    Plot.rectY(data, { x: "x", y: "y", interval: 0.05 })
  ]
})