UK Tricontours / Fil

Fil

Vocateur.

Workspace

Public

Edited

Dec 23

3 stars

display(contour.thresholds(40), triangles.flat())

contour = d3

.tricontour()

.triangulate(triangulate)

.thresholds(10)

triangulated = triangulate(data, d => d[0], d => d[1])

triangulate = (points, x, y) => {

const halfedges = Int32Array.from({ length: 3 * triangles.length }, () => -1);

const inedges = Int32Array.from({ length: points.length }).fill(-1);

const hull = [];

const index = new Map();

for (let i = 0; i < triangles.length; i++) {

const tri = triangles[i];

index.set([tri[0], tri[1]].join("/"), 3 * i);

index.set([tri[1], tri[2]].join("/"), 3 * i + 1);

index.set([tri[2], tri[0]].join("/"), 3 * i + 2);

}

let t = 0,

for (const tri of triangles) {

for (const e of [[tri[0], tri[1]], [tri[1], tri[2]], [tri[2], tri[0]]]) {

if ((a = index.get([e[1], e[0]].join("/"))) !== undefined) {

halfedges[a] = t;

halfedges[t] = a;

}

t++;

}

const _triangles = triangles.flat(2);

const segments = new Map(

Array.from(halfedges)

.map((j, i) =>

j === -1

? [_triangles[i], _triangles[i % 3 === 2 ? i - 2 : i + 1]]

: null

)

.filter(d => !!d)

);

let i, j, e;

while ((e = [...segments].pop())) {

i = e[0];

do {

j = i;

hull.push(j);

i = segments.get(j);

segments.delete(j);

} while (!hull.includes(i));

}

t = 0;

for (const tri of triangles) {

for (const e of [[tri[0], tri[1]], [tri[1], tri[2]], [tri[2], tri[0]]]) {

if (inedges[e[1]] === -1 || halfedges[t] === -1) {

inedges[e[1]] = t;

}

t++;

}

return {

points: points.map(d => [x(d), y(d)]).flat(),

triangles: _triangles,

halfedges,

inedges,

hull

};

}

md`

${display(d3.tricontour(), d3.Delaunay.from(data).triangles)}</div>

${display(contour.thresholds(10), triangles.flat())}</div>

d3 = require("d3@7", "d3-delaunay@6", "d3-tricontour@1")

scales = ({

x: d3.scaleLinear().range(d3.extent(data, d => d[0])),

y: d3.scaleLinear().range(d3.extent(data, d => d[1])),

color: d3

.scaleSequential(d3.interpolateCividis)

.domain(d3.extent(data, d => d[2]))

})

function display(contour, triangles) {

const svg = d3

.create("svg")

.style("background", "#333")

.attr("viewBox", [

scales.x(-1.5 / 20),

scales.y(-1.5 / 20),

scales.x(1 + 1.5 / 20) - scales.x(-1.5 / 20),

scales.y(1 + 1.5 / 20) - scales.y(-1.5 / 20)

]);

const g = svg.append("g");

const path = d3.geoPath().pointRadius(scales.x(2 / width) - scales.x(0));

const contours = contour(data);

g.append("g")

.selectAll("path")

.data(contours)

.join("path")

.attr("d", path)

.attr("fill", (d) => scales.color(d.value))

.attr("fill-opacity", 0.8)

.append("title")

.text((d) => d.value);

function vertices(a) {

return Array.from(triangles)

.filter((_, i) => i % 3 === a)

.map((d) => data[d]);

}

g.append("path")

.attr("vector-effect", "non-scaling-stroke")

.attr(

"d",

path({

type: "MultiLineString",

coordinates: d3.zip(vertices(0), vertices(1), vertices(2), vertices(0))

})

)

.attr("fill", "none")

.attr("stroke-width", 1)

.attr("stroke", "white");

const pts = g

.append("g")

.selectAll("circle")

.data(data)

.join("path")

.attr("d", (d) => path({ type: "Point", coordinates: d }))

.attr("fill", (d) => scales.color(d[2]));

path.pointRadius(2 * path.pointRadius());

const coordinates = Array.from(triangulated.hull).map((i) => data[i]);

coordinates.push(coordinates[0]);

g.append("path")

.attr("vector-effect", "non-scaling-stroke")

.attr(

"d",

path({

type: "LineString",

coordinates

})

)

.attr("fill", "none")

.attr("stroke-width", 2)

.attr("stroke", "red");

g.append("g")

.selectAll("circle")

.data(triangulated.hull)

.join("path")

.attr("d", (d) => path({ type: "Point", coordinates: data[d] }))

.attr("fill", "red");

g.append("g")

.selectAll("text")

.data(data)

.join("text")

.style("fill", "white")

.style("font-size", scales.x(14 / width) - scales.x(0))

.attr("transform", (d) => `translate(${d[0]}, ${d[1]})`)

.text((d, i) => i);

return svg.node();

}

One platform to build and deploy the best data apps

Experiment and prototype by building visualizations in live JavaScript notebooks. Collaborate with your team and decide which concepts to build out.

Use Observable Framework to build data apps locally. Use data loaders to build in any language or library, including Python, SQL, and R.

Seamlessly deploy to Observable. Test before you ship, use automatic deploy-on-commit, and ensure your projects are always up-to-date.

Learn more