Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add native range charts #7288

Open
ndrezn opened this issue Nov 26, 2024 · 2 comments
Open

add native range charts #7288

ndrezn opened this issue Nov 26, 2024 · 2 comments
Labels
feature something new P3 backlog

Comments

@ndrezn
Copy link
Member

ndrezn commented Nov 26, 2024

MinutePhysics called us out (by name!) for not having native support for range charts: https://www.youtube.com/watch?v=5zBg9pH_6bE

However, to be clear, it is possible to build this chart type in Plotly using the base attribute for bar charts and doing a bit of data preparation.

Here's a live demo of a custom range chart and our built-ins: https://codepen.io/ndrezn/pen/XJrWboB

Code snippet
/**
 * Creates a range bar chart using Plotly.js.
 *
 * @param {Object} data - The dataset, with keys as categories (e.g., cities) and values as arrays of measurements.
 * @param {Array} labels - The labels for each measurement (e.g., temperature metrics).
 * @param {string} containerId - The ID of the HTML container to render the chart.
 * @param {string} title - The title of the chart.
 */
function rangeChart(data, labels, containerId, title) {
  const categories = Object.keys(data); // Extract categories (e.g., cities)

  // Generate bar traces for each label
  const barTraces = labels.map((label, index) => {
    const bases = categories.map((category) =>
      index === 0 ? data[category][index] : data[category][index - 1]
    ); // base, i.e. the previous value
    const heights = categories.map(
      (category) => data[category][index] - bases[categories.indexOf(category)]
    ); // Height relative to the base, i.e. the current value
    return {
      x: categories,
      y: heights, // Bar heights
      base: bases, // Starting points for each bar
      name: label,
      type: "bar"
    };
  });

  // Layout for the chart
  const layout = {
    title: title,
    barmode: "overlay", // Overlapping bars to show ranges
    xaxis: { title: "Category" },
    yaxis: { title: "Value" }
  };

  // Render the chart
  Plotly.newPlot(containerId, barTraces, layout);
}

const ranges = {
  Ontario: [-9, 3, 8, 13, 27],
  England: [3, 8, 12, 16, 24],
  Kentucky: [-3, 8, 14, 20, 30]
};

const labels = [
  "Winter mean low",
  "Annual mean low",
  "Annual mean",
  "Annual mean high",
  "Summer mean high"
];

rangeChart(
  ranges,
  labels,
  "rangePlot",
  "Temperature Ranges in Three Londons (range)"
);

Image

That being said, having this chart-type built in would be great. The mental gymnastics are a bit tricky to get a true range chart working correctly.

Here are the built-in ways to achieve a similar chart:

Code snippet
// Data for the three Londons
const ranges = {
  Ontario: [-9, 3, 8, 13, 27],
  England: [3, 8, 12, 16, 24],
  Kentucky: [-3, 8, 14, 20, 30]
};

const labels = [
  "Winter mean low",
  "Annual mean low",
  "Annual mean",
  "Annual mean high",
  "Summer mean high"
];
const cities = Object.keys(ranges);

// Scatter plot traces
const scatterTraces = labels.map((label, index) => ({
  x: cities,
  y: cities.map((city) => ranges[city][index]),
  mode: "markers",
  name: label,
  type: "scatter"
}));

const scatterLayout = {
  title: "Temperature Ranges in Three Londons (scatter)",
  xaxis: { title: "City" },
  yaxis: { title: "Temperature" }
};

Plotly.newPlot("scatterPlot", scatterTraces, scatterLayout);

// Bar plot traces
const barTraces = labels.map((label, index) => ({
  x: cities,
  y: cities.map((city) => ranges[city][index]),
  name: label,
  type: "bar"
}));

const barLayout = {
  title: "Temperature Ranges in Three Londons (overlay)",
  barmode: "overlay",
  xaxis: { title: "City" },
  yaxis: { title: "Temperature" }
};

Plotly.newPlot("barPlot", barTraces, barLayout);

// Bar plot traces
const stackedBarTraces = labels.map((label, index) => ({
  x: cities,
  y: cities.map((city) => ranges[city][index]),
  name: label,
  type: "bar"
}));

const stackedBarLayout = {
  title: "Temperature Ranges in Three Londons (stacked)",
  barmode: "stack",
  xaxis: { title: "City" },
  yaxis: { title: "Temperature" }
};

Plotly.newPlot("stackedBarPlot", stackedBarTraces, stackedBarLayout);

However, these aren't quite right.

  • The scatter plot is most legible, but doesn't depict a range
  • The overlaid bar shows the right data, but because the base of all positive points is 0, they actually overlap on top of each other. We want the base to he previous, and for England, we want the base to be 3, not 0.
  • The stacked bar chart looks right at first glance, but it's actually summing all the temperatures (and of course, the base for England is at 0, not 3. And the negative values stack a bit weirdly.

Image

@ndrezn ndrezn changed the title Feature Request: Range Charts Feature Request: Native Range Charts Nov 26, 2024
@emilykl
Copy link
Contributor

emilykl commented Nov 26, 2024

Probably wouldn't be too hard to add this as a Plotly Express chart, px.range or similar.

@gvwilson gvwilson changed the title Feature Request: Native Range Charts add native range charts Nov 26, 2024
@gvwilson gvwilson added feature something new P3 backlog labels Nov 26, 2024
@ndrezn
Copy link
Member Author

ndrezn commented Nov 26, 2024

Yeah that's a good idea! It doesn't need to live upstream as a native chart in Plotly.js.

That being said it could be implemented not as a new chart type but as an option on Bar, e.g. end parameter or mode or something...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
feature something new P3 backlog
Projects
None yet
Development

No branches or pull requests

3 participants