Heatmap Tool
Visualize matrix data using colour intensity across rows and columns.
Chart Data
Note: What Is a Heatmap?
What It Is
A heatmap is a grid of colored cells where each cell's color encodes the value of one variable at the intersection of two dimensions, usually a row and a column. Because the eye registers color patterns before conscious attention, a large grid of hundreds of cells can be scanned in an instant, revealing hotspots, cold spots, clusters, and gradients that a table of numbers would hide. The two dimensions are most often categorical, such as hours of the day, days of the week, products, regions, or samples.
There are two families that share the name. The data heatmap colors each cell by a measured value, which is the form used in activity grids and correlation matrices. The geographic heat map instead colors regions on a map by density or intensity, such as crime, traffic, or population. Both use the same core idea: color intensity as a fast, glanceable encoding of quantity.
Common Use Cases
- Website activity by hour of day and day of week, showing exact peaks and lulls
- Correlation matrices showing how closely pairs of variables move together
- User engagement across app features and user segments
- Sales by product category and region
- Server error counts by endpoint and time bucket
- Gene expression levels across samples, a classic biology application
When to Use It vs Alternatives
Use a heatmap when the data forms a two-way grid and the message is in the pattern of highs and lows. If only one dimension changes over time, a line chart gives far more precision. If the grid is tiny, fewer than a dozen cells, a plain table may be just as clear and more exact. For geographic density, use a choropleth or dot density map rather than a data grid heatmap. And when exact values matter, pair the heatmap with numbers printed in the cells or shown in tooltips.
How to Use It Well
- Choose perceptually uniform color scales and avoid rainbow ramps that distort reading
- Keep the direction consistent: decide whether dark means high or low and never flip it
- Print the exact values in important cells when precision matters
- Normalize values so the scale is comparable, or normalize per row when comparing patterns rather than magnitudes
- Order rows and columns by value or by clustering so real structure is not hidden by arbitrary order
- Use a diverging scale with a neutral midpoint when the data contains both positive and negative values
- Add tooltips that reveal exact values on hover in interactive versions
Fun Facts
- Gene expression heatmaps became famous in the 1990s, and the clustered heatmap with a dendrogram is now a standard figure in biology papers
- Financial software pioneer Cormac Kinney trademarked the term "Heat Map" in the 1990s for a two-dimensional display of market data
- Modern statistical software and spreadsheets can produce a correlation heatmap from a data table with a single command
- Heatmaps exploit pre-attentive processing: the brain detects color patterns before conscious attention engages
- Weather services use heatmaps of temperature anomalies to show how a region departs from its seasonal average
- Chess and game analysis tools show opening popularity and move frequency as heatmaps over the board
- The same idea powers data-as-art projects that turn thousands of measurements into striking color fields