214 lines
5.2 KiB
TypeScript
214 lines
5.2 KiB
TypeScript
import * as _Plot from "npm:@observablehq/plot";
|
|
import { DOMParser, SVGElement } from "npm:linkedom";
|
|
import * as _vega from "npm:vega";
|
|
import * as _lite from "npm:vega-lite";
|
|
|
|
const defaultPlotSettings = {
|
|
grid: true,
|
|
margin: 50,
|
|
style: {
|
|
backgroundColor: "#fff",
|
|
},
|
|
};
|
|
|
|
/**
|
|
* Configure default plot settings
|
|
* @param options
|
|
*/
|
|
export function configurePlots(options: any) {
|
|
Object.assign(defaultPlotSettings, options);
|
|
}
|
|
|
|
export const document = new DOMParser().parseFromString(
|
|
`<!DOCTYPE html><html></html>`,
|
|
"text/html",
|
|
);
|
|
|
|
export const Plot = _Plot;
|
|
export const vega = _vega;
|
|
export const vegalite = _lite;
|
|
|
|
/**
|
|
* Draw side-by-side plots
|
|
* Example:
|
|
* ```ts
|
|
* const plt = sideBySidePlot({
|
|
* x: ['enginesize', 'horsepower', 'citympg', 'highwaympg'],
|
|
* y: ['price'],
|
|
* marks: [
|
|
* (x, y) => Plot.dot(data, {x, y, fill: "species"}),
|
|
* (x, y) => Plot.linearRegressionY(data, {x, y, stroke: "red"}),
|
|
* ],
|
|
* cols: 2
|
|
* })
|
|
* @param x List of x-axis targets
|
|
* @param y List of y-axis targets
|
|
* @param marks List of plot callbacks
|
|
* @param cols Number of columns
|
|
*/
|
|
export function sideBySidePlot(opts: {
|
|
x: string[];
|
|
y: string[];
|
|
marks: any[];
|
|
cols: number;
|
|
options;
|
|
}) {
|
|
const imgTags: string[] = [];
|
|
for (const xTarget of opts.x) {
|
|
for (const yTarget of opts.y) {
|
|
const plt = Plot.plot({
|
|
...(opts.options ?? defaultPlotSettings),
|
|
marks: opts.marks.map((fn) => fn(xTarget, yTarget)),
|
|
document,
|
|
});
|
|
plt.setAttribute("xmlns", "http://www.w3.org/2000/svg");
|
|
const svgUrl = `data:image/svg+xml;base64,${
|
|
btoa(unescape(encodeURIComponent(plt)))
|
|
}`;
|
|
imgTags.push(`<img title="${xTarget} / ${yTarget}" src='${svgUrl}'>`);
|
|
}
|
|
}
|
|
const output = `
|
|
<section style="display:grid;grid-template-columns: repeat(${
|
|
opts.cols ?? 2
|
|
}, 1fr);">
|
|
${imgTags.join("")}
|
|
</section>
|
|
`;
|
|
return {
|
|
[Symbol.for("Jupyter.display")]: () => ({
|
|
"text/html": output,
|
|
}),
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Histogram plot
|
|
*
|
|
* @param data
|
|
* @param x
|
|
* @param opts
|
|
* @returns Plot
|
|
*/
|
|
export function histPlot(
|
|
data: any[],
|
|
x = "column",
|
|
opts = { options: null, fn: "proportion" },
|
|
) {
|
|
return Plot.plot({
|
|
...(opts.options ?? defaultPlotSettings),
|
|
y: { grid: true },
|
|
marks: [
|
|
Plot.ruleY([0]),
|
|
Plot.ruleX([0]),
|
|
Plot.rectY(data, Plot.binX({ y: opts.fn ?? "count" }, { x: x })),
|
|
],
|
|
document,
|
|
});
|
|
}
|
|
|
|
export function oneBoxPlot(
|
|
data: any[],
|
|
y = "column",
|
|
opts = { options: null, box: null },
|
|
) {
|
|
return Plot.plot({
|
|
...(opts.options ?? defaultPlotSettings),
|
|
y: { grid: true },
|
|
marks: [
|
|
Plot.ruleY([0]),
|
|
Plot.boxY(data, { y, ...(opts.box ?? {}) }),
|
|
],
|
|
document,
|
|
});
|
|
}
|
|
|
|
export async function quantilePlotSVG(
|
|
data: any[],
|
|
x = "column",
|
|
opts = { width: 500 },
|
|
) {
|
|
const spec = {
|
|
data: { values: data },
|
|
width: opts.width,
|
|
"transform": [
|
|
{
|
|
"quantile": "price",
|
|
"as": ["prob", "value"],
|
|
},
|
|
{
|
|
"calculate": "quantileNormal(datum.prob)",
|
|
"as": "norm",
|
|
},
|
|
],
|
|
"layer": [
|
|
{
|
|
"mark": { type: "circle", size: 80 },
|
|
"encoding": {
|
|
"x": {
|
|
"field": "norm",
|
|
"type": "quantitative",
|
|
"title": "Theoretical Quantiles→",
|
|
},
|
|
"y": {
|
|
"field": "value",
|
|
"type": "quantitative",
|
|
"title": "Ordered Values→",
|
|
},
|
|
},
|
|
},
|
|
{
|
|
mark: { type: "line", color: "red" },
|
|
transform: [
|
|
{
|
|
"regression": "value",
|
|
"on": "norm",
|
|
},
|
|
],
|
|
"encoding": {
|
|
"x": { "field": "norm", "type": "quantitative" },
|
|
"y": { "field": "value", "type": "quantitative" },
|
|
},
|
|
},
|
|
],
|
|
};
|
|
let vegaspec = vegalite.compile(spec).spec;
|
|
var view = new vega.View(vega.parse(vegaspec), { renderer: "none" });
|
|
|
|
return await view.toSVG();
|
|
}
|
|
|
|
export function quantilePlot(data: any[], x = "column", opts = { width: 500 }) {
|
|
return {
|
|
[Symbol.for("Jupyter.display")]: async () => ({
|
|
"text/html": await quantilePlotSVG(data, x, opts),
|
|
}),
|
|
};
|
|
}
|
|
|
|
const svgDataUrl = (plt) =>
|
|
`data:image/svg+xml;base64,${btoa(unescape(encodeURIComponent(plt)))}`;
|
|
|
|
export function threeChart(data: any[], x = "column", opts = { width: 800 }) {
|
|
const hist = histPlot(data, x);
|
|
hist.setAttribute("xmlns", "http://www.w3.org/2000/svg");
|
|
const box = oneBoxPlot(data, x);
|
|
box.setAttribute("xmlns", "http://www.w3.org/2000/svg");
|
|
const qq = quantilePlotSVG(data, x);
|
|
return {
|
|
[Symbol.for("Jupyter.display")]: async () => ({
|
|
"text/html": `
|
|
<section style="display:flex;flex-direction: column; gap: 1em;">
|
|
<div style="gap:0.5em; display:flex; flex-direction: row; border: 1px solid black;">
|
|
<img src="${svgDataUrl(hist)}" style="width: 100%;">
|
|
<img src="${svgDataUrl(box)}" style="width: 65%">
|
|
</div>
|
|
<div style="padding:1em; border: 1px solid black;">
|
|
<img src="${svgDataUrl(await qq)}" style="width: 100%">
|
|
</div>
|
|
</section>
|
|
`,
|
|
}),
|
|
};
|
|
}
|