Design
Chars
Illustrate complex data relations, show trends, patterns and summarize information
Types
- COMPARISON - show diff and similarities in data
- bar and column charts - represent simple nominal data
- more complex data can be shown via stacked variations
- DOs // start from zero, sort data, use single collor
- horizontal/vertical dot plot - represent data with possibility to add color/dot size measurement
- DON’T use dot plot for dense data sets
- bar and column charts - represent simple nominal data
- TRENDING - represent data change over time
- line chart - represent one or more(usually up to 4) data change trend over time
- DON’Ts // mix measurements, add to much values(lines), mix units
- line chart - represent one or more(usually up to 4) data change trend over time
- RELATIONSHIP - show relation/correlation between values
- scatter plot - more complex variation of dot plot, often with size/color difference in dots
- DOs // start from zero, highlight patterns via colors
- scatter plot - more complex variation of dot plot, often with size/color difference in dots
- COMPOSITIONS - show relative values, or values total makeup
- pie chars - highlight relative value(percentages) of some absolute value
- DOs // add 2-5 values, sort data clockwise
- treemap - show composition of hierarchical data via nested rectangles
- DOs // sort left-to-right + largest-smallest + top-bottom
- pie chars - highlight relative value(percentages) of some absolute value
- DISTRIBUTION - show volume distribution, anomalies and patterns
- heatmap - show density/concentration patterns for single value via color change
- DOs // use single color + different shades
- matrix chart - same as heatmap, but including more data via bubble size
- DOs // use single color + different shades
- heatmap - show density/concentration patterns for single value via color change
- METRIC - show progress
- flat gauge(like line loading) - show progress towards some metric
- polar gauge - show progress in circle
- can be draining or filling the circle
- ratings chart - similar to flat gauge, but have fixed steps and step count
- example:
- rating: |x x x 0 0|
- flat: |====---|
- example:
- DOs // add reference points(average, target etc), keep polar gauge big enough and use flat for smaller spaces
- LOCATION - show data in geographic map
- map - geo areas are highlighted via color
- bubble map - dot chart on map
- geo map - dot chart with different dot colors
- DOs // show specific regions and not all globe, use single color
- PIPLINE - show progress in stages and flow between stages
- funnel chart - represent stages, based on value of stage
- DOs // use related values, sort values
- waterfall chart - break trending plot into stages, color rectangle from starting to ending stage values and you will get
tradingwaterfall chart- DOs // use meaningful colors(increase, decrease, no change), show start-end of some period of time broken into stages
- origami chart - show not sequential data, with single measure
- sankey chart - show how value flow between two groups
- can show negative data too
- funnel chart - represent stages, based on value of stage
Color - important part of storytelling and user guidance
- color pallet must be accessible and big enough
- color pallet must work with light/dark modes via hue change
- DOs
- associate value and color across all charts
- pie chart must use different colors and not gradient, use minimalistic colors
- use color with cultural meaning, but remember that different cultures == different meanings
- remember about accessibility
- don’t overload interface(keep around 5 elements)