Types of plots
Visualization Cheatsheet — Chart Types
A concise reference for undergraduate students summarizing common chart types, when to use them, what they show, and quick best-practice tips.
Overview
Visualizations help turn data into insight. Choosing the right chart depends on: the question you want to answer (task), the types of variables you have (quantitative vs categorical vs temporal vs geographic), and your audience. Common roles for visualizations include:
- Showing change over time
- Showing part–to–whole composition
- Showing distributions
- Comparing values between groups
- Observing relationships between variables
- Mapping geographic data
Quick reference table
| Chart type |
Primary use |
Data type / encoding |
When to prefer it |
| Line chart |
Change over time |
Time (x) → numeric (y) |
Continuous trends, multiple series |
| Bar chart |
Compare categories |
Categorical → numeric |
Ranked comparisons, discrete categories |
| Grouped / stacked bar |
Compare subgroups or parts of whole |
Categorical × categorical |
Compare groups or show composition across categories |
| Histogram |
Distribution of one numeric variable |
Numeric → bins |
Explore shape, skew, outliers |
| Density plot |
Smoothed distribution |
Numeric |
When you want a smooth estimate |
| Box plot |
Summarize distribution |
Numeric by group |
Compare medians, spread, outliers |
| Violin plot |
Distribution + density by group |
Numeric by group |
Show distribution shape per group |
| Scatter plot |
Relationship between two variables |
Numeric × numeric |
Correlation, clusters, outliers |
| Bubble chart |
Scatter with third variable |
Numeric × numeric + size/color |
Add magnitude to scatter points |
| Heatmap |
Matrix or 2-D density |
Two categorical / binned numeric axes |
Show intensity across a grid |
| Choropleth / Cartogram |
Geographic patterns |
Region → numeric |
Map-based comparisons |
| Area chart |
Part-to-whole over time |
Time with stacking |
Emphasize totals and composition (use carefully) |
| Pie / Donut |
Simple part-to-whole |
Categorical proportions |
Only for few categories and when exact comparison not required |
| Treemap / Marimekko |
Hierarchical or composition |
Hierarchy → area |
Many subparts, space-limited views |
| Candlestick / OHLC |
Financial time series |
Time → open/high/low/close |
Price movements in trading data |
| Funnel chart |
Process conversion |
Ordered stages |
Conversion/drop-off visualization (marketing, UX) |
| Parallel coordinates |
Multivariate comparison |
Many numeric attributes |
Observe patterns across many variables |
| Dumbbell / Slope chart |
Two-point comparisons |
Category → two numeric points |
Show change between two timepoints or conditions |
| Bullet chart |
Compare to benchmark |
Numeric with target |
Compact KPI vs target display |
Short descriptions & classroom tips
Line chart
- What: Points connected by lines (time on x-axis).
- Use when: Visualizing trends, seasonality, and multiple series.
- Tip: Keep series to a readable number (≤ 6). Use small multiples when there are many series.
Bar chart (including grouped & stacked)
- What: Rectangles whose heights encode value.
- Use when: Comparing values across categories. Grouped bars compare subgroups; stacked bars show part–to–whole.
- Tip: Avoid stacked bars when precise comparison of subparts is needed; use grouped bars or small multiples.
Histogram & Density plot
- What: Histogram bins numeric data; density smooths it.
- Use when: Inspecting distribution shape, skew, modality.
- Tip: Show both histogram and density overlay in teaching exercises; discuss bin width effects.
Box plot & Violin plot
- What: Box shows median, IQR, whiskers; violin shows density shape.
- Use when: Comparing distributions across groups.
- Tip: Use box plots for concise comparison and violins when distribution shape matters.
Scatter & Bubble charts
- What: Plot points for pairs of values; bubble adds size (third var).
- Use when: Exploring correlations, clusters, heteroskedasticity.
- Tip: Avoid bubbles when size comparisons are critical (perceptual issues); consider log scales for skewed axes.
Heatmap
- What: Grid colored by value.
- Use when: Showing two-dimensional intensity (e.g., correlation matrices, 2-D histograms).
- Tip: Cluster rows/columns to reveal structure.
Choropleth & Cartogram
- What: Map regions colored (choropleth) or resized (cartogram) to encode values.
- Use when: Spatial comparisons (population, rates).
- Tip: Normalize by population for rates; beware of small-area visual bias.
Area & Stacked area
- What: Line chart with filled area under the curve; stacking shows composition.
- Use when: Emphasize totals and contributions over time.
- Tip: Stacked areas can hide trends for smaller series—use carefully.
Pie & Donut charts
- What: Circle divided into slices.
- Use when: Very small number of categories and audience prefers simple proportion visuals.
- Tip: Prefer bar charts for precise comparisons; avoid many slices.
Treemap & Marimekko
- What: Nested rectangles sized by value.
- Use when: Represent hierarchical composition or many categories in limited space.
- Tip: Order and labels matter—interactive hover helps.
Specialized charts (Candlestick, Funnel, Bullet, Parallel coordinates)
- What: Domain-specific; good to introduce as examples of targeted visualization.
- Use when: Financial time series (candlesticks), conversion flows (funnels), benchmark comparisons (bullet), and multivariate patterns (parallel coords).
Best-practice checklist (short)
- Use the simplest chart that answers the question.
- Label axes and include units. Provide a clear title and a short caption.
- Use perceptually effective encodings (position > length > angle > area > color hue).
- Limit categories and series shown at once; use small multiples for scale.
- Consider accessibility: colorblind-safe palettes, high contrast, descriptive alt text.
- Annotate important values or outliers; use direct labeling when possible.
- Avoid 3D effects unless they add genuine information.
Suggested classroom exercises
- Pick the right chart: Give the same dataset and three analysis questions; students pick and justify a chart for each question.
- Tufte reflection: Ask students to rework a messy dashboard to follow Tufte’s principles (minimize ink, maximize data-ink ratio).
- Perception lab: Compare bar vs pie vs donut for the same proportions and test which students estimate most accurately.
- Small multiples: Students create small-multiple line charts to compare dozens of time series.
Further reading / resources