This entry is part of the Chart Index, the reference library for the Chart Design Field Guide.

A time range chart shows daily values across a bounded run of dates. It uses the familiar weekly calendar structure of a calendar heatmap without requiring a full year, making short operational periods — a campaign, an incident window, a delivery quarter — easier to scan. Every day is a cell; the cells are stacked seven to a column, one column per week; colour carries the value.

The form exists because the calendar heatmap has a fixed frame. A year-long grid drawn for eleven weeks is mostly empty, its cells are tiny, and its month labels are wrong more often than they are right. The time range chart keeps only the part of the calendar that the reader's question is about, and lets the cells grow to fill the space. It is the chart for what happened, day by day, between these two dates?

What it is

Each day is a coloured cell, arranged into weeks and weekdays between a start and end date. Columns are weeks; rows are weekdays, Monday at the top in the house convention so the weekend sits together at the bottom. Colour carries the measure on a sequential scale; position carries the date and the weekly rhythm. Days before the start or after the end are simply absent, so the first and last columns may be ragged. Month names sit above the columns in which each month begins.

The hero chart above is seventeen weeks of support tickets, from 2 March to 28 June 2026, one cell per day. The weekly rhythm is the first thing the eye reads: five darker rows of weekdays and two pale rows of weekend. Then the exceptions. Thursday 23 April is the darkest cell on the chart — an outage that opened two hundred tickets — with a dark Friday after it. The three weeks from 4 May are a darker block across every weekday: the autumn campaign, which lifted the daily count by half. The pale cells sitting in otherwise dark rows are the public holidays: Good Friday, Easter Monday and the King's Birthday. The colour scale is clipped at ninety so the outage does not flatten everything else into the same pale tint. Randomise it and the counts jitter; the rhythm and the block survive.

When to use it

Time range charts are the right choice when:

  • The analysis covers several weeks or months rather than a full year.
  • Weekday patterns, gaps, and exceptional days matter.
  • Readers know the relevant start and end dates, such as a campaign, incident, or delivery window.
  • The data is daily counts or rates with a natural low-to-high reading.
  • The audience already reads calendar heatmaps — this is the same grammar, cropped.

When not to use it

  • Precise trends and rates of change matter. Colour cannot carry slope. Use a line chart.
  • The interval is only a few days — a dozen cells is a table — or spans many years, where the columns become hairlines.
  • Missing days cannot be distinguished clearly from zero values. If the source drops days with no activity, a zero and a gap look the same unless the pipeline fills them.
  • The value is not daily. Hourly data wants an hour × day heatmap; weekly data wants a bar chart.
The outage fortnight as a line
The same tickets for the two weeks around the outage, as a line. The spike's height can be read — 212 on the Thursday, 121 on the Friday — and the recovery slope is visible. What has gone is the weekday grid: the reader has to count back from the labels to know which days were weekends.

Design principles

Distinguish zero from missing

Use a defined colour for zero and a neutral empty state for missing data. Operational readers need to know whether nothing happened or nothing was recorded. The house renderer draws a day with no entry in the hairline grey it uses for days outside the range, and a day whose value is zero in the palest step of the scale — so the pipeline must emit zeros, not drop the rows.

Zeros and gaps in six weeks of deploys
Weekend cells carry a zero and are drawn in the palest teal: nothing was deployed. The five hairline-grey cells in the third week are not zeros — the deploy log for that week was lost — and the difference is the difference between a quiet week and a missing one.

Preserve weekday alignment

Consistent rows make recurring Monday-to-Friday patterns visible. Starting the week on Monday puts Saturday and Sunday together at the bottom of every column, so the weekend reads as one pale band; starting on Sunday splits it into the top and bottom rows and the eye has to reassemble it. Label weekdays when the chart is large enough — the house default ticks Monday, Wednesday and Friday.

Weeks starting on Sunday
The same seventeen weeks of tickets with Sunday as the first row. The weekend is now the top row and the bottom row, and the working week is the block between them. Correct, but the reader has to learn that the pale stripe at the top and the pale stripe at the bottom are the same thing.
Weeks starting on Monday
The house default. The weekend is a single pale band along the bottom of every column and the five working days are a solid block above it. The campaign weeks, the outage and the holidays all read against that block without any reassembly.

Use a sequential scale, and clip it to the ordinary range

Daily magnitude normally moves from low to high. A single-hue sequential palette supports that ordered reading. But a sequential scale is only as useful as its domain, and one extreme day can claim the whole of it. Clip the domain to the range ordinary days occupy and let the exceptional day saturate: it is still the darkest cell, and everything else keeps its contrast.

Domain set by the data: the outage owns the scale
With the scale running to 212, the outage is the only dark cell. Ordinary weekdays, at forty to seventy, all fall in the palest two steps, the weekend rows are indistinguishable from them, and the campaign block has disappeared.
Domain clipped at ninety
The same data with the colour domain capped at ninety. The outage and its aftermath saturate at the darkest step, as they should, and the rest of the chart regains its structure: weekday against weekend, campaign weeks against ordinary ones, holidays as pale cells in a dark row.

Annotate exceptional dates

Label outages, launches, deadlines, and other known events that explain unusual cells. The chart can show that Thursday 23 April was extraordinary; only a label can say why. Where the tooling has no per-cell annotation, the subtitle names the dates, as the hero's does, and the reader finds them.

Let the cells fill the width

The form's advantage over the calendar heatmap is cell size. Set the height so that seven rows of square cells fill the available width for the number of weeks in range — about fifty pixels a cell for seventeen weeks in a full-width frame — and the reader can see each day rather than squint at it. If the cells come out smaller than a fingertip, the range is too long for the form.

Forty weeks in the same frame
The hero's range extended to forty weeks. The cells are a quarter the size, the month labels are three times as dense, and the outage has become a dot. The information is all still there; the reader's ability to see it is not. Beyond about twenty weeks the calendar heatmap, with its smaller cells and year frame, is the better tool.

Anatomy

The composition of a time range chart
An anatomical guide

Columns represent weeks, rows represent weekdays, cells represent dates, colour encodes value, and blank cells mark dates outside the range or missing observations. The bounds are the design decision; everything else is the calendar.

  • Calendar heatmap — a longer calendar view, often covering a year, with smaller cells and a fixed frame.
  • Table heatmap — two-dimensional categorical comparison without calendar structure; the right form for hour × day.
  • Line chart — continuous trend and slope over time.
  • Bar chart — for the same days as a sequence of readable heights, when the period is short enough for the bars to have labels.

Reading list

  • Few, S. (2009). Now You See It. Visual patterns in time-oriented data.
  • Tufte, E. (1990). Envisioning Information. On small multiples and the density of calendar layouts.
  • Wilke, C. (2019). Fundamentals of Data Visualization. On sequential colour scales and the domain decision.