What chart library to use in React?

A quick comparison of the most popular React Charting frameworks

What chart library to use in React?

posted in javascript on  • 

The question “What chart library to use in React?” came about when I had to actually find one to introduce in a project. As suggested by a colleague, I created this blog post with the results of my research, maybe saving others from having to do the same.

Charting Libraries

recharts/recharts : A Redefined chart library built on d3

airbnb/visx : A collection of reusable low-level visualization components. Built on d3

plouc/nivo : Nivo provides supercharged React components to easily build dataviz apps, built on d3.

FormidableLabs/victory : A collection of components for building interactive data visualizations. Built on d3

uber/react-vis : A collection of components to render common data visualization charts. Built on d3

reactchartjs/react-chartjs-2 : React components for Chart.js, the most popular charting library


Many are built using the very versatile d3.js library:

d3/d3 : Bring data to life with SVG, Canvas and HTML. 📊📈🎉


If you already used Chart.js before, you may want to choose for the react-chartjs-2 wrapper:

chartjs/Chart.js : Simple HTML5 Charts using the canvas tag

Overview

Library Package size Weekly downloads Rendering type Pricing Documentation Setup time demo
Recharts 4.52 MB 1,278,432 Svg MIT Very good 15:10
Visx 440 kB 363,002 Svg MIT Very good 28:58
Nivo 2.81 MB 296,292 Svg / HTML / Canvas MIT Good 28:31
Victory 2.95 MB 192,203 Svg MIT Very good 11:10
React-vis 1.80 MB 155,768 Svg / Canvas MIT Ok 34:39
React-chartjs-2 4.88 MB 948,010 Canvas MIT Ok 35:11
Library Dependency Commits Issues Charts Storybook Docs
Recharts D3 2 days ago (2,254) 470 / 2220 11 WIP Examples
Visx D3 1 month ago (3,156) 117 / 608 20+ CodeSandbox Somewhat Interactive
Nivo D3 4 days ago (1,992) 81 / 1393 20 YES Very Interactive
Victory D3 last week (8,429) 240 / 1491 10 NO Gallery
React-vis D3 1 month ago (891) 306 / 511 10 Yes Gallery
React-chartjs-2 Chart.js 4 months ago (544) 52 / 665 15 CodeSandbox Gallery

Columns:
Commits: Last commit (total commits)
Issues: Open Issues / Closed Issues
Charts: Since many libraries offer composable components, the real amount of charts is practically infinite. The amount is an indication of how many are documented! 😃

Data as of 19/04/2023.

Test Data Structure

[
  { name: "Bob", age: 11, },
  { name: "Tom", age: 13, },
  { name: "Bruno", age: 56, },
  { name: "Kyoe", age: 33, },
  { name: "Borat", age: 333, },
  { name: "Sasha", age: 68, },
  { name: "Onur", age: 27 }
]

Installation and Usage

How “easy” is to get a Line Chart working for each libary?

Check out the Github source for working examples of these!

Recharts

Time: 15 minutes.

A smooth experience. Lots of stuff out-of-the-box with little or no configuration (ex: tooltip and legend).

Example

Rechart line chart

Code

npm install --save recharts
import { useSelector } from "react-redux";
import { CartesianGrid, Legend, Line, LineChart, Tooltip, XAxis, YAxis } from "recharts";

export function Recharts() {
  const persons = useSelector(state => state.persons);

  return (
    <LineChart
      width={500}
      height={300}
      data={persons}
      margin={{ top: 5, right: 30, left: 20, bottom: 5 }}
    >
      <CartesianGrid strokeDasharray="3 3" />
      <XAxis dataKey="name" />
      <YAxis />
      <Tooltip />
      <Legend />
      <Line type="monotone" dataKey="age" stroke="#8884d8" />
    </LineChart>
  );
}

Visx

Time: 29 minutes

The x and y axis configuration was done in a bit of a roundabout way, but the main problem and what took the longest time to set correctly was the tooltip. For some reason the library makes the tooltip data object use properties that have no relevance to the data being inserted in the chart. It assumes all data inserted are datum objects whereas in this case they are person objects. The documentation also neglects to mention this in their example.

Example

Visx line chart

Code

npm install --save @visx/xychart
import { Axis, Grid, LineSeries, Tooltip, XYChart } from "@visx/xychart";
import { useSelector } from "react-redux";

export function Visx() {
  const persons = useSelector(state => state.persons);

  const accessors = {
    xAccessor: (d ?: IPerson) => d && d.name,
    yAccessor: (d ?: IPerson) => d && d.age,
  };

  return (
    <XYChart
      width={500}
      height={300}
      xScale={{ type: "band" }}
      yScale={{ type: "linear" }}
    >
      <Axis orientation="bottom" />
      <Axis orientation="left" />
      <Grid columns={false} numTicks={4} />
      <LineSeries dataKey="Line 1" data={persons} {...accessors} />
      <Tooltip
        snapTooltipToDatumX
        snapTooltipToDatumY
        showVerticalCrosshair
        showSeriesGlyphs
        renderTooltip={({ tooltipData, colorScale }) => (
          <div>
            <div style={{ color: colorScale?.(tooltipData?.nearestDatum?.key!) }}>
              {tooltipData?.nearestDatum?.key}
            </div>
            {accessors.xAccessor(tooltipData?.nearestDatum?.datum)}
            {", "}
            {accessors.yAccessor(tooltipData?.nearestDatum?.datum)}
          </div>
        )}
      />
    </XYChart>
  )
}

Nivo

Time: 29 minutes

My main gripe and time sink with this package was the issue that the data properties needed to be called x and y respectively, so I needed to map my dataset to fit this mold. A simple setting for setting what properties you want to use on which axis would have been nice. Big visual issue however as you can probably tell is that the values don’t exactly show correctly on the x axis.

Example

Nivo line chart

Code

npm install --save @nivo/core @nivo/line
import { Line } from "@nivo/line";
import { useSelector } from "react-redux";

export function Nivo() {
  const persons = useSelector(state => state.persons);
  const data = persons.map(p => ({ x: p.name, y: p.age }));

  return (
    <Line
      width={500}
      height={300}
      curve="monotoneX"
      margin={{ top: 20, right: 20, bottom: 60, left: 80 }}
      enableSlices="x"
      data={[{ id: "persons", data: data }]}
      xScale={{ type: "band" }}
      yScale={{ type: "linear" }}
      axisLeft={{ legend: "linear scale", legendOffset: 12 }}
      axisBottom={{ legend: "linear scale", legendOffset: -12 }}
    />
  );
}

Victory

Time: 11 minutes

Usage was relatively easy, but it has the same limitation as the Nivo library because you also need to map your data properties to x and y. Also visually it doesn’t look as pleasing as some of the other charts in this list.

Example

Victory line chart

Code

npm install --save victory
import { useSelector } from "react-redux";
import { VictoryChart, VictoryContainer, VictoryLine } from "victory";

export function Victory() {
  const persons = useSelector(state => state.persons);
  const data = persons.map(p => ({ x: p.name, y: p.age }));

  return (
    <>
      <VictoryChart
        height={300}
        width={500}
        containerComponent={<VictoryContainer responsive={false} />}
      >
        <VictoryLine data={data} />
      </VictoryChart>
    </>
  );
}

React-vis

Time: 35 minutes

I could not get this chart to render remotely well. The documentation did not really help either in this case. Also I needed to map the data properties to x and y as well again.

Example

React-vis line chart

Code

npm install --save react-vis
import { useSelector } from "react-redux";
import { XYPlot, XAxis, YAxis, HorizontalGridLines, LineSeries } from "react-vis";

export function ReactVis() {
  const persons = useSelector(state => state.persons);
  const data = persons.map(p => ({ x: p.name, y: p.age }));

  return (
    <XYPlot width={500} height={300} xType="ordinal">
      <HorizontalGridLines />
      <XAxis />
      <YAxis />
      <LineSeries color='red' data={data} />
    </XYPlot>
  );
}

React-chartjs-2

Time: 35 minutes

Documentation for this library was kind of limited, especially when it came to creating a line chart. An annoying trait is that you have to split your dataset up in into a labels array for the x axis and data array for the y axis. The end result does look nice however.

Example

React-chartjs-2 line chart

Code

npm install --save react-chartjs-2 chart.js
import { useSelector } from "react-redux";
import { Chart as ChartJS, CategoryScale, LinearScale, PointElement, LineElement, Title, Tooltip, Legend } from "chart.js";
import { Line } from "react-chartjs-2";

ChartJS.register(CategoryScale, LinearScale, PointElement, LineElement, Title, Tooltip, Legend);

const options = {
  responsive: false,
  plugins: {
    legend: {
      position: "top" as const,
    },
    title: {
      display: true,
      text: "Chart.js Line Chart",
    },
  },
};

export function ReactChartJS2() {
  const persons = useSelector(state => state.persons);
  const labels: string[] = persons.map(person => person.name);
  const ages: number[] = persons.map(person => person.age);
  const data = {
    labels,
    datasets: [{
      label: "Dataset 1",
      data: ages,
      borderColor: "rgb(255, 99, 132)",
      backgroundColor: "rgba(255, 99, 132, 0.5)",
    }],
  };

  return (
    <Line width={500} height={300} options={options} data={data} />
  );
}

Conclusion

Onur: Recharts

First time was a charm, the first library Recharts did everything I wanted it to in a clear and convenient manner. It’s the only library where I didn’t really have a complaint. It’s easy to use and has very good documentation.

Wouter: Nivo

Disclaimer: I didn’t actually implement the Line Charts.

The out of the box experience of the Line Chart was, for me, really nice for Recharts, Nivo and React-ChartJS-2.

I was really impressed with the documentation of Nivo: a very interactive gallery plus a storybook implementation, just wow. The dev team also seems to be on top of things with 81 open issues and a whopping 1393 closed ones. It’s the only library with a below v1 version though and has not yet gained as much traction (300k weekly downloads) compared to current leaders Recharts (1.5M) and React-chartjs-2 (950k), the latter which coasts of the popularity of Chart.js itself.

Deployed Demo App

The create-react-app is published at itenium.be.

Demo App Overview


Stuff that came into being during the making of this post
Tags: react