A box plot summarizes a measure's full distribution in one shape: a box spanning the middle 50% of values (the interquartile range), a line at the median, whiskers extending to the most extreme non-outlier values, and individual points for outliers. It shows spread and skew at a glance, something an average or a single bar can't communicate.

How are outliers determined?

Using the Tukey method: the interquartile range (IQR) is Q3 minus Q1, and any value falling more than 1.5 times the IQR below Q1 or above Q3 is flagged as an outlier. Outliers are drawn as separate points and excluded from the whiskers, so the whisker ends always show the most extreme non-outlier values, not the absolute min and max.

Is the category field required?

No. Box Plot works with just a single measure and no dimension at all, showing one box for the entire pooled dataset. Assigning a categorical or date field splits that single box into one box per category, so you can compare distributions side by side. Removing the field is how you switch back to the pooled, ungrouped view.

Why is there no aggregation dropdown on the measure field?

A box plot needs every individual raw value to calculate quartiles correctly - averaging or summing first would destroy the distribution it's meant to show. The measure field always fetches raw, unaggregated values, so the usual aggregation dropdown is replaced with a fixed '(Default)' label, since there's no legitimate alternative to offer.

Can I change how outlier points look?

Yes. Boxplot Styles includes Show Outliers, Outlier Color, Outlier Symbol (circle, rectangle, rounded rectangle, triangle, diamond, or pin), Outlier Size, and Outlier Opacity, all independent of the box's own fill and border styling.

When should I use a box plot instead of a histogram?

Use a box plot when you want to compare a distribution's spread and outliers across multiple categories side by side - it's compact enough to show many groups in one chart. Use a histogram when you want the actual shape of a single distribution's frequency, which a box plot's five-number summary can't show (two very different-shaped distributions can produce nearly identical boxes).

Does the tooltip show the same thing when I hover the box versus an outlier point?

No. Hovering the box itself shows the full five-number summary - Category, Max, Q3, Median, Q1, and Min. Hovering an individual outlier point shows only Category and Value, since a single point doesn't have quartiles of its own.

Box Plot | Edilitics Docs

Show a measure's quartiles, median, and outliers in one shape, optionally split by category. Best for comparing spread and skew, not just averages.

A box plot summarizes a measure's full distribution - not just its average - using quartiles, a median line, and outlier points. Drop exactly one numeric field into Row. A categorical or date field in Column is optional: without it, you get one box for the entire dataset; with it, one box per category, side by side. Use it when comparing spread and outliers matters as much as comparing typical values.

When to Use

An average or a single bar tells you nothing about whether the underlying values are tightly clustered or wildly spread out, or whether a few extreme outliers are skewing the picture. A box plot answers exactly that: the box shows where the middle half of your data sits, the line inside it shows the median, the whiskers show the realistic range, and any points outside the whiskers are flagged as outliers worth investigating individually.

The category field is optional, and removing it is how you get the pooled view. A box plot with no dimension assigned still works - it shows a single box summarizing every value in the measure column. Assign a category to split that into one box per group for comparison; remove the category to go back to one pooled box. There's no "Default" aggregation choice to pick - dropping the field is the toggle.

Switch to a different chart when:

You want the actual shape of one distribution's frequency, not a five-number summary - use a Histogram
You need exact values for each record, not a statistical summary - use Text Table
You're comparing totals or counts across categories, not spread - use Horizontal Bar

Scenario	Dimension (optional)	Measure
Compare delivery time spread across carriers	Carrier	Delivery time (days)
Salary distribution across departments	Department	Salary
Response time spread for a single service	None	Response time (ms)
Order value spread by customer segment	Customer segment	Order value
Exam score spread across class sections	Class section	Score

Required Inputs

Field	Type	Count
Dimension	Categorical or Date	0 to 1
Measure	Numeric	Exactly 1

For step-by-step build instructions, see Build Your First Chart.

Formatting Options

The Format tab unlocks after the measure field is assigned.

Style

Use the chart title to state what's being compared, especially when several boxes sit side by side.

Control	What it does
Show Chart Title	Shows or hides the title.
Enter Chart Title	Title text. Maximum 50 characters.
Font family	Font applied to the title.
Font size	5 to 30.
Bold / Italic	Weight and style.
Alignment	Left, center, or right within the chart container.

Boxplot Styles is where the box's shape, color, and outlier points are controlled, all in one panel.

Control	What it does
Layout	Horizontal or vertical orientation of the boxes.
Fill Color	Color of the box body.
Border Color / Width / Type	Color, thickness (0 to 5), and line style (solid, dashed, dotted) of the box's outline, including the whisker lines.
Min Width / Max Width	The box's width bounds in pixels (1 to 50 for minimum, 10 to 100 for maximum) - ECharts scales actual width within this range based on how many boxes are shown and available space.
Show Outliers	Shows or hides outlier points entirely, without affecting the whisker calculation - whiskers always exclude outliers regardless of whether the points themselves are visible.
Outlier Color	Color of the outlier markers, independent of the box's fill color.
Outlier Symbol	Circle, rectangle, rounded rectangle, triangle, diamond, or pin.
Outlier Size	2 to 20px.
Outlier Opacity	0.1 to 1.0.

Axes

Axis Name labels each axis - the category field's name on one axis, the measure's name on the other, both editable.

Control	What it does
Show Axis Name	Shows or hides the axis title. Falls back to the assigned field's name in Title Case, or "Category" / "Value" when no field is assigned.
Custom label	Overrides the automatic field-name label for that axis.
Font Family / Color / Font size	Styling for the axis name text.

Axis Line controls the visibility and color of the line each axis sits on.

Control	What it does
Show Axis Line	Shows or hides the line itself.
Line Color	Color of the axis line.

Axis Label controls the text printed along each axis - category names on one side, the measure's value scale on the other.

Control	What it does
Show Axis Label	Shows or hides the labels.
Font Family / Color / Font size	Styling for the labels.
Text Case	Uppercase, lowercase, capitalize, or none, for category labels.

Category labels longer than 14 characters truncate automatically with an ellipsis, and rotate at an angle once more than 10 categories are shown, so a wide category set stays readable without manual adjustment.

GridLines controls the reference lines behind the boxes, set independently for the X-axis and Y-axis via the axis selector at the top of the panel.

Control	What it does
X Axis / Y Axis selector	Chooses which axis's grid lines the controls below apply to.
Show Gridlines	Shows or hides that axis's grid lines.
Grid Color	Color of that axis's grid lines, when shown.
Position (Left / Right / Top / Bottom)	Percentage-based margins for the chart's plot area within its container - shared across both axes, not per-axis like Show and Color above.

Interactivity

The tooltip's content depends on what you hover: a box shows the full five-number summary, while an outlier point shows just its own value.

Control	What it does
Show Tooltip	Shows or hides the tooltip entirely.
Header / value text styling	Font, size, and color for the category and value lines shown in the tooltip.

Hovering the box itself shows Category, Max, Q3, Median, Q1, and Min in one panel. Hovering an individual outlier point shows only Category and Value, since a single point has no quartiles of its own.

Animation controls the boxes' transition when the chart first renders or the data changes.

Control	What it does
Enable Animation	Turns the transition animation on or off.
Duration	How long the transition takes.
Delay	How long the transition waits before starting.
Easing Function	The transition curve.

Emphasis controls the visual response when hovering over a box - a scale-up effect and a drop shadow. Off by default.

Control	What it does
Show Emphasis	Turns hover effects on.
Focus Type	Item, Series, or None.
Enable Scale	Scales the hovered box up slightly.
Scale Size	1.0x to 2.0x.
Shadow Blur / Color / Offset X / Offset Y	Drop shadow styling on the hovered box.
Border Width	Border added around the hovered box.

Enable the Toolbox when viewers need to export the chart or inspect its underlying data.

Control	What it does
Show Toolbox	Shows or hides the toolbox icon bar.
Save as Image	Adds a download icon that saves the chart as a PNG.
Data View	Adds an icon that opens the underlying data table in a separate view.

Best Practices

Start without a category, then add one once you know what you're comparing. A single pooled box tells you the overall spread; assigning a category only makes sense once you have a specific comparison in mind - department against department, carrier against carrier. Adding a category you don't actually need just produces more boxes to scan.

Investigate outliers individually rather than dismissing them. The Tukey method's 1.5×IQR rule is a statistical convention, not a judgment that those points are wrong - an outlier in delivery time might be a genuine fulfillment failure worth following up on, not noise to filter out.

Don't compare box plots with very different sample sizes without noting it. A box built from 5 values and one built from 5,000 values can look equally tidy, but the larger sample's quartiles are far more reliable. Box plots don't show sample size on their own - mention it in the title or alongside the chart if it matters to the comparison.

Use Layout to match the orientation of your category labels. Horizontal layout (the default) works well for a handful of short category names; switching to Vertical can help when there are many categories or longer labels that would otherwise need heavy rotation.

Remember a box plot can hide multimodal distributions. Two clusters of values with a gap between them can still produce a single, ordinary-looking box. If a distribution's actual shape matters - not just its quartiles - pair the box plot with a Histogram of the same measure.

FAQs

Need help? Email support@edilitics.com with your workspace, job ID, and context. We reply within one business day.

Box Plot

What is a box plot?