How to use Recharts to visualize analytics data (with examples)

Aug 14, 2024

Posted by

On this page

Recharts is a popular charting library for React. It provides many visualization and customization options to use with analytics data in PostHog.

To provide examples of what you can do with it, we create a basic React app, set it up to get data from PostHog's query API, and visualize it with Recharts.

Don't want to do all the work of querying and visualizing? You can always share or embed insights, dashboards, and more. See sharing docs for more.

Creating a basic React app

To start, we will create a basic React app using Vite.

Terminal
npm create vite@latest rechart-ph -- --template react

Once created, go into the folder, install the dependencies, and start the dev server.

Terminal
cd rechart-ph
npm install
npm run dev

This gives us a basic React app we can add our example visualizations to.

Basic React app

Querying data from PostHog

Assuming you have data in PostHog to query, the next step is to set up our query request to PostHog.

Don't have data to query? Check out our guide to setting up analytics in React to start capturing some. Unlike other tutorials, we aren't going to install PostHog in this one.

Start by creating a personal API key. You can do this by going to personal API keys in your project settings, clicking Create personal API key, giving it a label, choosing the Performing analytics queries scope (AKA query read), and clicking Create key.

Creating a personal API key in PostHog

⚠️ Warning: The following is a simplified example. Exposing your personal API key (like we do below) exposes your private PostHog data to the public internet, so don't do this in real life. Make sure your personal API key isn't exposed and your query request happens securely on the server side.

With this personal API key, we can set up a request in our React app to PostHog's query API endpoint using your project ID. We use HogQL in the payload to get the data we want and display it unformatted in our app for now:

JavaScript
import { useState, useEffect } from 'react'
import './App.css'


function App() {
  const [data, setData] = useState([])
  const projectId = "12345"


  useEffect(() => {
    const fetchQuery = async () => {
      const url = `https://us.posthog.com/api/projects/${projectId}/query/`;
      const headers = {
        "Content-Type": "application/json",
        "Authorization": "Bearer phx_1a2b3c4d5e6f7g8h9i0j1k2l3m4n5o6p7q8r9s0tuvwxyz"
      };
  
      const payload = {
        "query": {
          "kind": "HogQLQuery",
          "query": "select properties.$current_url from events where properties.$current_url != null limit 10"
        }
      }
      try {
        const response = await fetch(url, {
          method: "POST",
          headers: headers,
          body: JSON.stringify(payload),
        });
        const data = await response.json();
        setData(data.results)
      } catch (error) {
        console.error("Error fetching data:", error);
      }
    }
    fetchQuery();
  }, []);


  return (
    <>
      <pre>{JSON.stringify(data, null, 2)}</pre>
    </>
  )
}


export default App

The page we end up with is just an ugly collection of non-aggregated data, but we'll fix that next.

Ugly data

Visualizing data with Recharts

Now that we have our app and data source set up, it is time to visualize it. To start, we install Recharts:

Terminal
npm install recharts

Once done, we can simply import the necessary components along with our query. Some useful queries and visualizations include:

Line chart of pageviews by day

A basic query to start with is getting the count of $pageview events by day. To do this, we:

  1. Write a query to select the date and count() of $pageview events grouped and ordered by date.
  2. Change the format of the data from the array PostHog provides the object Recharts expects.
  3. Import and use the LineChart, Line, XAxis, YAxis, and Tooltip components from recharts to create our chart.

Together, this looks like this:

JavaScript
import { useState, useEffect } from 'react'
import './App.css'
import { LineChart, Line, XAxis, YAxis, Tooltip } from 'recharts';


function App() {
  const [data, setData] = useState([])
  const projectId = "12345"


  useEffect(() => {
    const fetchQuery = async () => {
      const url = `https://us.posthog.com/api/projects/${projectId}/query/`;
      const headers = {
        "Content-Type": "application/json",
        "Authorization": "Bearer phx_1a2b3c4d5e6f7g8h9i0j1k2l3m4n5o6p7q8r9s0tuvwxyz"
      };
  
      const payload = {
        "query": {
          "kind": "HogQLQuery",
          "query": `SELECT 
                      toDate(timestamp) AS date,
                      count() AS pageview_count
                    FROM events
                    WHERE event = '$pageview'
                    GROUP BY date
                    ORDER BY date DESC
                    LIMIT 20`
        }
      }
      try {
        const response = await fetch(url, {
          method: "POST",
          headers: headers,
          body: JSON.stringify(payload),
        });
        const data = await response.json();
        setData(data.results.reverse());
      } catch (error) {
        console.error("Error fetching data:", error);
      }
    }
    fetchQuery();
  }, []);


  const formattedData = data.map(([date, count]) => ({
    date,
    pageviews: count
  }));


  return (
    <>
      <LineChart width={600} height={400} data={formattedData}>
        <XAxis dataKey="date" />
        <YAxis />
        <Tooltip />
        <Line type="monotone" dataKey="pageviews" stroke="#8884d8" />
      </LineChart>
    </>
  )
}


export default App

This creates a basic chart of pageviews by date:

Line chart of pageviews by day

Bar chart of most popular paths

We can write a similar query to get the top paths by a count of pageviews. We basically swap in properties.$pathname AKA pathname, filter out null values, and order by the pageview_count.

The challenge is getting all this data to fit. To do this, we set the Rechart BarChart component to the vertical layout and increase the width of both the whole component and the YAxis. On top of this, we change the colors to make it stand out.

Together this looks like this:

JavaScript
//... imports
import { BarChart, Bar, XAxis, YAxis, Tooltip } from 'recharts';


function App() {


    // ... useState, useEffect, fetch request, etc.
  
      const payload = {
        "query": {
          "kind": "HogQLQuery",
          "query": `SELECT 
                      properties.$pathname AS pathname,
                      count() AS pageview_count
                    FROM events
                    WHERE 
                        event = '$pageview' 
                        AND properties.$pathname IS NOT NULL 
                        AND properties.$pathname != '/'
                    GROUP BY pathname
                    ORDER BY pageview_count DESC
                    LIMIT 10`
        }
      }
      try {
        const response = await fetch(url, {
          method: "POST",
          headers: headers,
          body: JSON.stringify(payload),
        });
        const data = await response.json();
        setData(data.results);
      } catch (error) {
        console.error("Error fetching data:", error);
      }
    }
    fetchQuery();
  }, []);


  const formattedData = data.map(([pathname, count]) => ({
    pathname,
    pageviews: count
  }));


  return (
    <>
      <BarChart width={1000} height={400} data={formattedData} layout='vertical' >
        <XAxis type="number" stroke='white' />
        <YAxis type="category" dataKey="pathname" width={300} stroke='white' />
        <Tooltip />
        <Bar dataKey="pageviews" fill="#FFA500" activeBar={false} />
      </BarChart>
    </>
  )
}


export default App

This creates a bar chart that looks like this:

Bar chart of most popular paths

Stacked area of browsers

We can do a more complicated visualization of pageviews by browser. The tricky part is formatting the data to work with the visualization. The input data from PostHog looks like this:

JavaScript
[
  ['2023-05-01', 'Chrome', 100],
  ['2023-05-01', 'Firefox', 50],
  ['2023-05-02', 'Chrome', 120],
]

The data we want for the visualization looks like this:

JavaScript
[
  { date: '2023-05-01', Chrome: 100, Firefox: 50 },
  { date: '2023-05-02', Chrome: 120, Firefox: 0 }
]

This means we must loop through the data, create entries for each date, and add pageviews for each browser to the date entry. Once formatted, we can use the AreaChart to visualize it. Together, this looks like this:

JavaScript
//... imports
import { AreaChart, Area, XAxis, YAxis, Tooltip } from 'recharts';


function App() {
            // ... useState, useEffect, fetch request, etc.
  
      const payload = {
        "query": {
          "kind": "HogQLQuery",
          "query": `SELECT 
                      toDate(timestamp) AS date,
                      properties.$browser AS browser,
                      count() AS pageview_count
                    FROM events
                    WHERE event = '$pageview' AND properties.$browser IS NOT NULL
                    GROUP BY date, browser
                    ORDER BY date DESC, pageview_count DESC
                    LIMIT 20`
        }
      }
      try {
        const response = await fetch(url, {
          method: "POST",
          headers: headers,
          body: JSON.stringify(payload),
        });
        const data = await response.json();
        console.log(data.results);
        setData(data.results.reverse());
      } catch (error) {
        console.error("Error fetching data:", error);
      }
    }
    fetchQuery();
  }, []);


  const browsers = [...new Set(data.map(([_, browser]) => browser).filter(Boolean))];


  const formattedData = data.reduce((acc, [date, browser, count]) => {
    const existingDate = acc.find(item => item.date === date);
    if (existingDate) {
      existingDate[browser] = count;
    } else {
      const newEntry = { date };
      browsers.forEach(b => newEntry[b] = 0);
      newEntry[browser] = count;
      acc.push(newEntry);
    }
    return acc;
  }, []);


  const colors = ['#8884d8', '#82ca9d', '#ffc658', '#ff7300', '#a4de6c', '#d0ed57', '#83a6ed', '#8dd1e1'];


  return (
    <>
      <AreaChart width={1000} height={400} data={formattedData}>
        <XAxis dataKey="date" />
        <YAxis />
        <Tooltip />
        {browsers.map((browser, index) => (
          <Area 
            key={browser}
            type="monotone"
            dataKey={browser}
            stackId="1"
            stroke={colors[index % colors.length]}
            fill={colors[index % colors.length]}
          />
        ))}
      </AreaChart>
    </>
  )
}


export default App

This gives a visualization that looks like this:

Stacked area chart of browsers

Pie chart of operating systems

The last visualization we can set up is a pie chart of operating systems. We can get the data by getting a count of $pageview events by properties.$os. We then format this data and set up custom labels and colors for the pie chart.

Together, this looks like this:

JavaScript
//... imports
import { PieChart, Pie, Cell } from 'recharts';


function App() {
            // ... useState, useEffect, fetch request, etc.
  
      const payload = {
        "query": {
          "kind": "HogQLQuery",
          "query": `SELECT 
                      properties.$os AS os,
                      count() AS pageview_count
                    FROM events
                    WHERE event = '$pageview'
                      AND properties.$os IS NOT NULL
                    GROUP BY os
                    ORDER BY pageview_count DESC`
        }
      }
      try {
        const response = await fetch(url, {
          method: "POST",
          headers: headers,
          body: JSON.stringify(payload),
        });
        const data = await response.json();
        console.log(data.results);
        setData(data.results);
      } catch (error) {
        console.error("Error fetching data:", error);
      }
    }
    fetchQuery();
  }, []);


  const formattedData = data.map(([os, pageview_count]) => ({
    name: os,
    pageview_count
  }));


  const colors = ['#0088FE', '#00C49F', '#FFBB28', '#FF8042', '#8884D8'];


  return (
    <>
      <PieChart width={600} height={400}>
        <Pie
          data={formattedData}
          dataKey="pageview_count"
          nameKey="name"
          cx="50%"
          cy="50%"
          label={({ name, percent }) => `${name} ${(percent * 100).toFixed(0)}%`}
          labelLine={false}
        >
          {
            formattedData.map((entry, index) => (
              <Cell key={`cell-${index}`} fill={colors[index % colors.length]} />
            ))
          }
        </Pie>
      </PieChart>
    </>
  )
}


export default App

This creates a nice-looking pie chart with labels like this:

Pie chart of operating systems

Further reading

Comments