Does not work with typed array data columns

[etpinard]

Consider:

const regl = require('regl')({ extensions: 'oes_element_index_uint' })
const createMatrix = require('./')

let splom = createMatrix(regl)

splom.update({
  data: [
    new Float32Array([1, 2, 3]),
    new Float32Array([4, 6, 8]),
    new Float32Array([1, 5, 9])
  ],
  ranges: [
    [0, 10],
    [0, 10],
    [0, 10]
  ],
  domain: [
    [0, 0, 0.3, 0.3],
    [0.35, 0.35, 0.65, 0.65],
    [0.7, 0.7, 1, 1] 
  ],
  color: ['red', 'green', 'blue'],
  size: [20, 10, 30]
})

splom.draw()

this gives:

and a blank screen.

---

Is this a known limitation of this implementation? That is, should we try to convert typed array to plain arrays in plotly.js? Or would there be an easy way to make regl-splom accept typed arrays?

[dy]

Should be fixed in 1.0.1.

[etpinard]

Thanks @dy ! Sploms are now working with Float32Array, but they are still broken with Int32Array (as in plotly/plotly.js#2585) and possibly other types.

For example,

const regl = require('regl')({ extensions: 'oes_element_index_uint' })
const createMatrix = require('./')

let splom = createMatrix(regl)

splom.update({
  data: [
    new Int32Array([1, 2, 3]),
    new Int32Array([4, 6, 8]),
    new Int32Array([1, 5, 9])
  ],
  ranges: [
    [0, 10],
    [0, 10],
    [0, 10]
  ],
  domain: [
    [0, 0, 0.3, 0.3],
    [0.35, 0.35, 0.65, 0.65],
    [0.7, 0.7, 1, 1] 
  ],
  color: ['red', 'green', 'blue'],
  size: [20, 10, 30]
})

splom.draw()

gives

and a blank screen.

---

Looks like regl-scatter2d accepts Int32Array just fine. Could we make regl-splom do the same?

[dy]

Ok, now should be fixed )

[etpinard]

Sweet. This works:

lotly.newPlot( gd,
			  [{
    "type": "splom",
    "dimensions": [{
      "values": new Int32Array([1, 2, 3]),
      "label": "A"
    }, {
      "values": new Int32Array([2, 5, 6]),
      "label": "B"
    }]
  }]);

gives