StackGenVis: Alignment of Data, Algorithms, and Models for Stacking Ensemble Learning Using Performance Metrics
https://doi.org/10.1109/TVCG.2020.3030352
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
84 lines
3.2 KiB
84 lines
3.2 KiB
4 years ago
|
# regl-splom
|
||
|
|
|
||
|
|
Matrix of scatter plots. A wrapper over [regl-scatter2d](https://github.com/dy/regl-scatter2d) for optimized intersected data rendering.
|
||
|
|
|
||
|
|
* [x] minimal GPU memory footprint: N vs N*N in direct [regl-scatter2d](https://github.com/dy/regl-scatter2d) passes case.
|
||
|
|
* [ ] optimized performance due to binary trees for 1d point clustering, opposed to default 2d quad clustering.
|
||
|
|
|
||
|
|
|
||
|
|
## Usage
|
||
|
|
|
||
|
|
[](https://npmjs.org/package/regl-splom/)
|
||
|
|
|
||
|
|
```js
|
||
|
|
let regl = require('regl')({extensions: 'oes_element_index_uint'})
|
||
|
|
let createMatrix = require('regl-splom')
|
||
|
|
|
||
|
|
let scatterMatrix = createMatrix(regl)
|
||
|
|
|
||
|
|
// pass data and views to display
|
||
|
|
scatterMatrix.update(
|
||
|
|
{ data: [[], [], ...], ranges, domains, viewport, size, color, border },
|
||
|
|
{ data: [[], [], ...], ranges, domains, viewport, size, color, border }
|
||
|
|
)
|
||
|
|
|
||
|
|
// draw views by ids
|
||
|
|
scatterMatrix.draw(0, 1, ...views)
|
||
|
|
```
|
||
|
|
|
||
|
|
## API
|
||
|
|
|
||
|
|
### `splom = createSplom(regl)`
|
||
|
|
|
||
|
|
Creates scatter matrix instance.
|
||
|
|
|
||
|
|
### `splom.update(trace1, trace2, ...traces)`
|
||
|
|
|
||
|
|
Define passes for `draw` method. Every trace can include the following options:
|
||
|
|
|
||
|
|
Option | Description
|
||
|
|
---|---
|
||
|
|
`data` | An array with arrays for the columns.
|
||
|
|
`range` | Array with data ranges corresponding to `data`. Every range can be an array `[min, max]` or `[minX, minY, maxX, maxY]`. If undefined - detected automatically.
|
||
|
|
`domain` | Array with domains for the data, ie. the area data dimension holds within the `viewport`. Each domain can be an array `[min, max]` for symmetric placement or `[minX, minY, maxX, maxY]` for precise position. Domain values are from `0..1` interval, defining what area of the `viewport` a dimension holds. By default domains cover viewport evnely.
|
||
|
|
`padding` | Padding within domains (in px), or list of paddings per-domain. By default `[0,0,0,0]`. Can be a number, an array or any [rectangle](https://github.com/dy/parse-rect) format.
|
||
|
|
`color`, `size`, `borderColor`, `borderSize`, `opacity` | Points style.
|
||
|
|
`marker` | Points marker.
|
||
|
|
`diagonal` | Show or hide diagonal.
|
||
|
|
`upper` | Show or hide upper half matrix.
|
||
|
|
`lower` | Show or hide lower half matrix.
|
||
|
|
`viewport` | Area that the plot holds within the canvas. Can take any [rectangle](https://github.com/dy/parse-rect) format.
|
||
|
|
<!-- `transpose` | Use transposed view of data, ie. swap columns and rows. -->
|
||
|
|
<!-- `normalizeDomain` | Normalize domains to fit the viewport. -->
|
||
|
|
<!-- `snap` | Enable snapping for the points, ie. hide invisible points -->
|
||
|
|
|
||
|
|
|
||
|
|
### `splom.draw(...ids?|...points?)`
|
||
|
|
|
||
|
|
Draw all defined passes, or only selected ones provided by `ids`. Optionally define point indexes to render.
|
||
|
|
|
||
|
|
```js
|
||
|
|
// draw 1 and 3 passes
|
||
|
|
splom.draw(1, 3)
|
||
|
|
|
||
|
|
// draw 1, 2 and 3 points from the first pass and 3 point from the second pass
|
||
|
|
splom.draw([1, 2, 3], [3])
|
||
|
|
```
|
||
|
|
|
||
|
|
### `splom.destroy()`
|
||
|
|
|
||
|
|
Dispose renderer and all the associated resources
|
||
|
|
|
||
|
|
## Related
|
||
|
|
|
||
|
|
* [regl-scatter2d](https://github.com/dy/regl-scatter2d)
|
||
|
|
* [regl-line2d](https://github.com/dy/regl-line2d)
|
||
|
|
* [regl-error2d](https://github.com/dy/regl-error2d)
|
||
|
|
|
||
|
|
|
||
|
|
## License
|
||
|
|
|
||
|
|
© 2018 Dmitry Yv. MIT License
|
||
|
|
|
||
|
|
Development supported by [plot.ly](https://github.com/plotly/).
|