Deep Scatterplots for the Web

This is a library for allowing interactive visualization of extremely large datasets, in browser.

It's fast for three reasons:

1. All data is sent in the Apache Arrow feather format, in a custom quadtree format that makes it possible to only load data as needed on zoom. Feather takes no time to parse in the browser once transferred, compresses pretty well, and can be directly copied to the GPU without transformation in JS. This is the way of the future. We also now allow custom in-browser calculations on Arrow data.
2. Most rendering is done in custom layers using WebGL, with a buffer management strategy handled by REGL. This means that there are no unnecessary abstractions around points or separate draw calls for different objects; a minimum number of buffers are attached for all needed points.
3. Almost all grammar-of-graphics transforms such are handled on the GPU, which allows for interpolated transitions with calculations done in parallel.

It also runs in completely static settings, so you can host a million-point scatterplot over something like Github Pages.

Deepscatter is the underlying graphics engine for Atlas. It is provided under an NC-CC-BY-SA license for all noncommercial use.

Examples

• Atlas map of 5.5 million tweets. (Nomic)
• Narrative visualization of 20 million biomedical abstracts embedded using T-SNE.. (Collaboration between Nomic and the University of Tübingen).
• 1 million+ documents from arxiv.com rendered inside an Observable notebook. (Ben Schmidt)
• Every person in the 2010 and 2020 US Censuses displayed in an interactive svelte-kit app. (Ben Schmidt)
• Newspaper Articles at the Library of Congress from the Reconstruction Era. (Andromeda Yelton while in residency at the Library of Congress).

Documentation

See the autogenerated typescript documentation for a full list of methods and functions.

Get help

Github issues, even low quality ones, are welcome here.

There is also a dedicated Deepscatter Slack which you are welcome to join. Nomic maintains a discord server that has dedicated Deepscatter and Atlas support channels.

I came into doing this stuff from a very non-technical background and welcome people to join with naive questions.

Quick start

Importing the module.

See the arxiv example above to see some basic examples.

Running locally.

First, install the companion tiling library, which is written in python, and generate a million points of test data in tiles of 50000 apiece.

python3 -V # requires Python 3.9.x or 3.10.x
python3 -m pip install git+https://github.com/bmschmidt/quadfeather
quadfeather-test-data 1_000_000
quadfeather --files tmp.csv --tile_size 50_000 --destination tiles

Then setup this library to run. It will start a local dev server.

npm i
npm run dev

If you go to localhost:3344, you should see an interactive scatterplot. To dig into what you're seeing, open index.html. (In 2021, this development site works in Chrome, not Safari or Firefox, because it uses ES6 module syntax inside the webworker. The distributed version of the module should work in all browsers.)

Your own data.

1. Create a CSV, parquet, or feather file that has columns called 'x' and 'y'. (Or a feather file that has columns x, y). Any other columns (categorical information, etc.) can be included as additional columns.

2. Tile it:

cd deepscatter # if you're not already there
quadfeather --files ../some-path-to/your-data.csv --tile_size 50000 --destination tiles

3. Assuming your dataset has an x and y column and the tiles folder is in the root directory of this project, you can see the data visualized by running

npm run dev

and opening http://localhost:3345/index-simplest-way-to-start.html in your browser.

To edit the visualization, or to troubleshoot, look at the file index-simplest-way-to-start.html, where you should find a bare-bones implementation of deepscatter.

Explore index.html, and render it at http://localhost:3345/index.html, for a more advanced example.

Note: Ideally, in a future release you'll be able to create these specs in away that doesn't require coding JSON directly.

Build the module

npm run build

will create an ES module at dist/deepscatter.es.js The mechanics of importing this are very slightly different than index.html.

Note that this is an ESM module and so requires you to use <script type="module"> in your code. Don't worry! We're allowed to do this now! But do be aware that this will not work on computers running very old browsers.

Snippet:

<div id="my-div"></div>
<script type="module">
  import Scatterplot from './dist/deepscatter.umd.js';
  f = new Scatterplot('#my-div');
</script>

See index_prod.html for an example

This is currently bundled with vite and rollup. There is/will be a further interaction layer on top of it, but the core plotting components are separate and should work as a standalone layer that supports plot requests via an API.

Code strategy

Any interaction logic that changes the API call directly does not belong in this library. The only interaction code here is for zooming and interacting with points.

Future codebase splits.

The plotting components and the tiling components are logically quite separate; I may break the tiling strategy into a separate JS library called 'quadfeather'.

Apache Arrow would still be a necessary intermediate format, but it could be generated from CSV files using, say, arquero or a WASM port of DuckDB.

API

This is still subject to change and is not fully documented. The encoding portion of the API mimics Vega-Lite with some minor distinctions to avoid deeply-nested queries and to add animation and jitter parameters.

{
   encoding: {
     "x": {
         "field": "x",
         "transform": "literal"
     },
     "color": {
         "field": "year",
         "range": "viridis",
         "domain": [1970, 2020]
   }
}

Implemented aesthetics.

1. x
2. y
3. size
4. jitter_radius: size of jitter. API subject to change.
5. jitter_speed: speed of jitter. API subject to change.
6. color (categorical or linear: range can call color scales explicitly, or accepting any d3-color name.)
7. x0 (for animations; transitions between x0 and x)
8. y0 (for animations; transitions between y0 and y)
9. filter. (Filtering is treated as an aesthetic operation by this library.)

Planned

1. Symbol (Mapping of categorical variables to single unicode points in a single font; probably 255 max.)
2. Label (Full-text label)
3. Image (Like PixPlot)

Jitter

Jitter is a little overloaded with features right now, but some are quite fun.

jitter method is set on 'method' key of the 'jitter_radius' field. Possible values are:

1. circle
2. spiral
3. time
4. normal

Principles

1. This is a 2d library. No fake 3d.
2. The central zoom state is handled by d3-zoom.
3. Use the zoom state to render other layers on top of Deepscatter by hooking in (note on_zoom is directly set, not passed in via prefs):

const scatterplot = new Scatterplot('#deepscatter');
scatterplot.on_zoom = (transform) => {...}