快速浏览这里的内容，点击任意模板打开。
A heatmap answers one question well: where's it dense, and where's it empty, across two categorical axes at once. Plot hour of day against day of week and you can spot exactly when a gym fills up (Monday at 6pm, Wednesday at 6am) in one glance that a bar chart would need fifty-six separate bars to match. Or plot signup cohort against month and you get a retention grid that shows churn patterns a line chart would blur into a tangle of overlapping series.
PlotSet currently ships one heatmap template, built around a straightforward category-by-category matrix: a row category, a column category, and a numeric value in each cell. That single template covers most real use cases, from attendance patterns to correlation matrices to survey score grids, because the data shape is flexible even though the chart type itself isn't. What actually changes from case to case is the color scale, not the chart: sequential for a plain count that only goes up from zero, diverging when values can fall above or below a baseline.
Color choice matters more here than in most chart types, because color is the entire encoding. Get it wrong and the chart misleads. Perceptually uniform scales, where equal steps in the data produce equal-looking steps in color, are the safer default; rainbow scales exaggerate some ranges and flatten others, which is why most modern data-viz style guides have moved away from them. Around 8% of men have some form of red-green color blindness, so a red-to-green diverging scale is a real accessibility problem, not a nitpick.
每种都有专属的免费制作工具，且已加载示例数据。
Foot traffic, gym check-ins, support ticket volume: anything that repeats on a daily and weekly cycle. Twenty-four hours across seven days gives you 168 cells, small enough to read as one shape, and the peaks and dead zones jump out immediately.
Signup month down one axis, months-since-signup across the other, retention percentage in each cell. It's the standard SaaS retention view because the diagonal decay pattern is instantly recognizable once you've seen it once.
Every variable plotted against every other variable, color showing correlation strength from -1 to 1. Exam scores across subjects, stock returns across sectors: anywhere you're hunting for which pairs move together.
Product feature adoption by customer segment, or survey response by demographic group. When there are too many category combinations for a bar chart to stay readable, a grid still works because you're reading color, not comparing bar lengths.
从电子表格到已发布的图表，一步一步来。
Two category columns (row category and column category) plus one numeric value column, one row per cell. PlotSet pivots that into a grid automatically, and a wide format with one column per category works too.
It loads with sample data bound in, so you can see the exact column layout it expects before you touch your own file.
Import a CSV or Excel file, then choose sequential or diverging depending on whether your values only run from zero or cross a meaningful midpoint.
Publish for an interactive version with hover tooltips showing the exact value per cell, or export a static PNG, SVG or PDF for a report or deck.
打开编辑器，从该类别的任何模板开始（已加载示例数据）。
打开编辑器A heatmap shows density across two categorical axes, like hour and weekday or cohort and month. A choropleth map shows one value per geographic region, shaded onto actual map shapes. If both of your axes are categories, use a heatmap; if one of them is literally geography, use a map template instead.
免费开始，300+ 模板任选。无需信用卡。
