dataviz

name	dataviz
description	Data visualization advisor. Recommends chart types based on data shape, audience, and goal. Covers 150+ visualization types with taxonomy from datavizproject.com. Triggers on "what chart should I use", "visualize this data", "dataviz", "chart type", "which visualization", "how to show/display/plot".
argument-hint	[data description or question]

Data Visualization Advisor

Recommend the right visualization for any dataset, explain chart types, and guide implementation decisions.

Based on: datavizproject.com by Ferdio — the world's largest data visualization library (160+ types).

Use this for: Choosing chart types, understanding visualization trade-offs, mapping data shape to visual form, exploring alternatives to common charts.

---

How to Use This Skill

When invoked, follow this decision process:

Step 1: Understand the Data

Ask (or infer from context):

1. What are you showing? — The subject/story
2. How many variables? — 1, 2, 3, many
3. What type of variables? — Categorical, continuous, temporal, geographic, hierarchical, network
4. How many data points? — Handful, dozens, hundreds, thousands+
5. Who is the audience? — General public, analysts, executives, domain experts

Step 2: Identify the Function

Map the user's goal to one of these seven functions:

Function	The user wants to...	Primary chart families
Comparison	Compare values across categories or groups	Bar, Grouped Bar, Lollipop, Dot, Bullet, Radar
Trend over time	Show how values change over a time period	Line, Area, Stream, Sparkline, Slope, Bump, Candlestick
Part to whole	Show how parts contribute to a total	Pie, Donut, Treemap, Sunburst, Waffle, Marimekko, Stacked Bar
Distribution	Display the spread or shape of data	Histogram, Box Plot, Violin, Beeswarm, Strip, Jitter
Correlation	Reveal relationships between variables	Scatter, Bubble, Connected Scatter, Heat Map, Parallel Coordinates
Geolocation	Show data tied to geographic places	Choropleth, Dot Density, Bubble Map, Flow Map, Cartogram
Concept visualization	Explain a process, structure, or idea	Flow Chart, Sankey, Venn, Mind Map, Org Chart, Illustration

Step 3: Recommend (with Reasoning)

Provide:

1. Primary recommendation — The best-fit chart type with rationale
2. Runner-up — A strong alternative and when to prefer it
3. Avoid — A common but poor choice for this case and why
4. Implementation notes — Library suggestions (D3, Observable Plot, Vega-Lite, Chart.js, matplotlib, ggplot2, etc.)

---

Taxonomy System

Every visualization is classified along four orthogonal dimensions:

Family (what kind of visual)

Family	Description
Chart	Traditional statistical/quantitative graphics (bar, line, area, pie)
Diagram	Conceptual, relational, or process-oriented visuals (flow, Venn, org chart)
Geospatial	Map-based visualizations (choropleth, dot map, flow map)
Plot	Statistical/mathematical coordinate-based visuals (scatter, box, violin)
Table	Tabular/grid data arrangements (heat map, matrix, comparison chart)

Function (what purpose it serves)

Comparison · Concept visualization · Correlation · Distribution · Geolocation · Part to whole · Trend over time

Shape (visual form)

Area · Circle/Circular · Line · Point · Rectangle/Bar · Square · Grid · Arc/Radial · Spiral

Input (data requirements)

One variable · Two variables · Three variables · Several variables · Temporal · Hierarchical · Network/relational · Geographic · Text

---

Complete Visualization Catalog

Bar & Column Charts

Type	When to use	Key trait
Bar Chart	Compare values across categories	The workhorse — use by default for comparison
Bar Chart (Horizontal)	Many categories or long labels	Easier to read category names
Grouped Bar Chart	Compare 2-3 series across categories	Side-by-side bars per group
Stacked Bar Chart	Part-to-whole AND comparison	Segments stack; use 100% variant for pure proportions
Butterfly Chart	Compare two opposing groups (male/female, before/after)	Also called Tornado Chart
Triangle Bar Chart	Stylistic variant of bar chart	Triangular bars
Lollipop Chart	Cleaner alternative to bar chart	Line + dot; less ink, same data
Bullet Graph	Show value against a target/threshold	Replaces gauges on dashboards
Span Chart	Show ranges (min-max) per category	Two Y values per point
Column Range	Show floating ranges per category	Also called floating column chart
Waterfall Chart	Show cumulative positive/negative changes	Running total — great for financials
Pareto Chart	Highlight the vital few factors	Bars (descending) + cumulative line
Histogram	Show distribution of continuous data	Bins, not categories — distinct from bar chart
Radial Bar Chart	Comparison with circular layout	Bar chart on polar coordinates
Radial Histogram	Distribution of cyclical data	Histogram wrapped around a circle
Spiral Histogram	Compare cycles while keeping timeline	Also called condegram
Pictorial Bar Chart	Add visual metaphor to bars	Icons replace/overlay bars

Line & Trend Charts

Type	When to use	Key trait
Line Graph	Continuous data over time	Connect the dots — the default for time series
Spline Graph	Smoother time series	Fitted curve instead of straight segments
Stepped Line Graph	Data that changes at discrete points	Step function — good for pricing tiers, rates
Slope Chart	Compare exactly two time points	Two-point before/after comparison
Bump Chart	Show rank changes over time	Focus on position, not magnitude
Sparkline	Inline trend within text or table	Word-sized, no axes — pure trend
Radial Line Graph	Cyclical patterns (hours, months)	Data wrapped around a circle
Connected Scatter Plot	Two variables evolving over time	Time is encoded in the path direction
Kagi Chart	Price movements ignoring time	Only draws when price moves enough
Candlestick Chart	OHLC financial data	Open/high/low/close per period
Renko Chart	Price trend filtering noise	Equal-size bricks, direction changes on threshold
Control Chart	Statistical process control	Line + upper/lower control limits

Area Charts

Type	When to use	Key trait
Area Chart	Emphasize volume/magnitude over time	Filled line chart
Stacked Area Chart	Part-to-whole over time	Multiple series stacked vertically
Layered Area Chart	Compare overlapping series	Transparency-based layering
Radial Area Chart	Cyclical magnitude data	Area chart on polar coordinates
Stream Graph	Organic, flowing multi-series over time	Displaced around central axis
Sorted Stream Graph	Same but sorted for readability	Series sorted by magnitude
3D Stream Graph	Surface-style stream	Three-dimensional variant

Pie & Donut Charts

Type	When to use	Key trait
Pie Chart	2-5 clearly different proportions	Controversial but effective for simple part-to-whole
Donut Chart	Same as pie, with space for a metric in the center	Functionally identical to pie
Semi Circle Donut	Part-to-whole with visual economy	Half donut saves space
Multi-level Donut	Hierarchical proportions	Concentric rings for parent-child
Polar Area Chart	Compare magnitudes across equal-angle sectors	Nightingale Rose / Coxcomb chart
Sunburst Diagram	Multi-level hierarchical part-to-whole	Concentric rings from root outward
Compound Bubble & Pie	Geographic proportional composition	Bubble size + pie composition
Pie Chart on a Map	Proportions at geographic locations	Pie charts placed on map
Fan Chart	Past trend + future uncertainty	Line graph meets range/prediction area

Scatter & Bubble Plots

Type	When to use	Key trait
Scatter Plot	Relationship between two continuous variables	The default for correlation
Bubble Chart	Three continuous variables	x, y, + bubble size
Circular Bubble Chart	Three variables on circular plane	Alternative layout for bubble
3D Scatter Plot	Three continuous variables in 3D	Use sparingly — hard to read in 2D print
Hexagonal Binning	Scatter with thousands of overlapping points	Density via hexagonal bins

Dot & Strip Charts

Type	When to use	Key trait
Dot Chart	Simpler alternative to bar chart	Cleaner for few values
Dot Plot	Simple 1D distribution	Dots on a number line
Strip Plot	Distribution of one continuous variable	Single-axis scatter
Jitter Plot	Strip plot with overlap reduction	Random offset to reveal density
Beeswarm Plot	Distribution with zero overlap	Systematic positioning, not random
Dumbbell Plot	Compare two values per category	Two dots connected by a line

Statistical Distribution

Type	When to use	Key trait
Box Plot	Summarize distribution (median, quartiles, outliers)	The statistician's default
Violin Plot	Full distribution shape + box plot	Shows density, not just quartiles
Error Bars	Show uncertainty/precision of measurements	Added to other chart types

Proportional Area

Type	When to use	Key trait
Proportional Area (Circle)	Quick size comparison, no axes	Circle area encodes value
Proportional Area (Square)	Same but with squares	Easier to compare than circles
Proportional Area (Half Circle)	Compact proportional comparison	Half-circle variant
Proportional Area (Icon)	Add meaning via icon shape	Scaled icons
Nested Proportional Area	Layered proportional comparison	Nested/overlapping shapes
Packed Circle Chart	Hierarchical data as tangent circles	Circle packing layout

Pictorial / Icon-Based

Type	When to use	Key trait
Pictorial Unit Chart	Make data tangible for general audience	Icons instead of abstract shapes
Pictorial Fraction Chart	"1 in 5 people..." style	Filled vs. unfilled icons
Pictorial Percentage Chart	Percentage via pictogram fill	Icons showing percentage
Pictorial Stacked Chart	Part-to-whole with visual metaphor	Stacked pictogram bars
Icon Count	Simple count visualization	Repeated icons
Icon and Number	Hero stat with visual context	Number + illustrative icon

Numbers & Simple Indicators

Type	When to use	Key trait
Scaled-up Number	Single standout metric (KPI)	Big bold number — no chart needed
Progress Bar	Show completion toward a goal	Linear fill
Waffle Chart	Percentage with discrete units	10×10 grid, colored cells = percentage
Tally Chart	Frequency count with raw feel	Tally marks
Angular Gauge	Show value within a range	Speedometer-style dial
Solid Gauge	Good/bad indicator	Index gauge (above/below threshold)

Heat Maps & Matrices

Type	When to use	Key trait
Heat Map	Dense 2D data as color intensity	Grid of colored cells
Bubble-based Heat Map	Heat map + magnitude via bubble size	Circles in a grid
Circular Heat Map	Cyclical 2D data (hours × days)	Radial layout
Spiral Heat Map	Compare continuous cycles	Spiral preserves timeline + cyclicality
Matrix Diagram	Present/absent relationships	Intersection grid
Comparison Chart	Feature/attribute comparison	Rows × columns table
Table Chart	Precise values matter more than shape	Just a table — sometimes the best choice

Network & Relationship Diagrams

Type	When to use	Key trait
Network Visualization	Complex many-to-many relationships	Nodes and edges (force-directed)
Chord Diagram	Flows between entities in a closed set	Arcs around a circle
Non-ribbon Chord Diagram	Same, emphasis on connections not volume	Lines instead of ribbons
Arc Diagram	Connections with ordered nodes	1D layout, arcs above/below
Sankey Diagram	Flow quantities through stages	Width = quantity; energy/cost transfers
Alluvial Diagram	Category changes over time	Flows between parallel categorical axes
Hive Plot	Network with structural axes	Nodes mapped to radial axes by property
Radial Convergences	Entity relationships in strict circle	Circular node layout with connections
Clustered Force Layout	Hierarchical grouping with nested circles	Circle packing + force simulation
Parallel Coordinates	High-dimensional multivariate data	Vertical axes, one per variable
Parallel Sets	Categorical flow/frequency	Like parallel coordinates for categories

Hierarchical / Tree Visualizations

Type	When to use	Key trait
Treemap	Part-to-whole in hierarchical data	Nested rectangles, area = value
Convex Treemap	Same with non-rectangular polygons	Aesthetic alternative
Dendrogram	Cluster hierarchy from analysis	Tree branching from clustering
Organisational Chart	Reporting structure	Classic top-down tree
Mind Map	Brainstorm / idea relationships	Central concept + branches
Hyperbolic Tree	Very large hierarchies with focus+context	Fisheye distortion for exploration
Fan Chart (Genealogy)	Family tree	Half-circle with concentric ancestor rings

Pyramid & Funnel

Type	When to use	Key trait
Funnel Chart	Sequential filtering/drop-off	Progressively narrowing stages
Pyramid Chart	Inverted funnel / hierarchical tiers	Broadest at base
Pyramid Diagram	Conceptual hierarchy (Maslow's etc.)	Conceptual, not data-driven
Population Pyramid	Age/gender distribution	Back-to-back horizontal bars

Flow & Process Diagrams

Type	When to use	Key trait
Flow Chart	Decision logic / process steps	Boxes + arrows
Cycle Diagram	Repeating process with no clear start/end	Circular flow
Step by Step Illustration	Sequential instructions	Illustrated panels
Illustration Diagram	Explain how something works	Annotated illustration
Development and Causes	Timeline + explanation of changes	Compound: timeline + area/line
SWOT Analysis	Strategic assessment	Four-quadrant grid

Venn & Set Diagrams

Type	When to use	Key trait
Venn Diagram	All logical relations between sets	Overlapping circles
Euler Diagram	Set relationships (may not show all combos)	Simpler than Venn when sets don't all overlap
Marimekko Chart	Two-dimensional part-to-whole	Variable-width stacked columns

Geographic / Map Visualizations

Type	When to use	Key trait
Choropleth Map	Values by region (density, rate, index)	Shaded regions — the default geo viz
Dot Density Map	Show concentration/presence	Many dots = high density
Bubble Map	Values at specific locations	Sized bubbles on map
Connection Map	Routes/links between locations	Lines connecting points on map
Flow Map	Directional movement between places	Width = quantity moved
Pin Map	Specific locations with labels	Pins/markers on map
Isoline Map	Continuous surface data (elevation, weather)	Contour lines on map
Cartogram	Distort geography to encode data	Area = data variable
Topographic Map	Physical terrain with elevation	Contour map
Profile Map	Elevation cross-section	Exaggerated vertical scale

Contour & Ternary Plots

Type	When to use	Key trait
Contour Plot	3-variable relationship on 2D surface	Like a topographic map of data
Ternary Plot	Three variables that sum to constant	Equilateral triangle coordinates
Ternary Contour Plot	Mixture response surface	Contours on ternary triangle

Radar & Multivariate

Type	When to use	Key trait
Radar Diagram	Compare multivariate profiles	Spider/web chart — axes from center
Taylor Diagram	Model-observation comparison	Correlation + RMS + amplitude in one plot

Timelines

Type	When to use	Key trait
Timeline	Chronological events (sequence)	Events in order
Scaled Timeline	Events with true time spacing	Proportional time distances
Bubble Timeline	Events with a magnitude variable	Bubble size on timeline
Gantt Chart	Project schedule / task durations	Horizontal bars across time

Word & Text

Type	When to use	Key trait
Word Cloud	Word frequency in text corpus	Size = frequency/importance

---

Quick Decision Flowchart

What do you want to show?
│
├─ COMPARISON across categories
│  ├─ Few categories (≤7) → Bar Chart
│  ├─ Many categories (8+) → Horizontal Bar Chart
│  ├─ Two groups side-by-side → Grouped Bar / Butterfly Chart
│  ├─ With a target/benchmark → Bullet Graph
│  └─ Multivariate profile → Radar Diagram
│
├─ CHANGE OVER TIME
│  ├─ Single series → Line Graph
│  ├─ Multiple series (≤4) → Multi-line Graph
│  ├─ Multiple series (5+) → Small multiples / Sparklines
│  ├─ Part-to-whole over time → Stacked Area / Stream Graph
│  ├─ Two time points → Slope Chart
│  ├─ Rank changes → Bump Chart
│  └─ Financial OHLC → Candlestick / Kagi / Renko
│
├─ PART TO WHOLE
│  ├─ 2-5 parts → Pie / Donut Chart
│  ├─ Many parts → Treemap
│  ├─ Hierarchical → Sunburst / Multi-level Donut
│  ├─ Exact percentages matter → Stacked Bar (100%) / Waffle
│  ├─ Sequential stages → Funnel Chart
│  └─ Two categorical dimensions → Marimekko Chart
│
├─ DISTRIBUTION
│  ├─ Single variable → Histogram / Strip Plot
│  ├─ Compare groups → Box Plot / Violin Plot
│  ├─ Show every point → Beeswarm / Jitter Plot
│  └─ 2D density → Hexagonal Binning / 2D Histogram
│
├─ CORRELATION / RELATIONSHIP
│  ├─ Two variables → Scatter Plot
│  ├─ Three variables → Bubble Chart
│  ├─ Many variables → Parallel Coordinates / Heat Map
│  ├─ Flow between entities → Sankey / Chord Diagram
│  ├─ Network structure → Network Graph / Arc Diagram
│  └─ Set overlap → Venn / Euler Diagram
│
├─ GEOGRAPHIC
│  ├─ Values by region → Choropleth Map
│  ├─ Points/locations → Pin Map / Dot Density Map
│  ├─ Values at locations → Bubble Map
│  ├─ Movement/flow → Flow Map / Connection Map
│  └─ Distort for data → Cartogram
│
├─ CONCEPT / PROCESS
│  ├─ Decision logic → Flow Chart
│  ├─ Hierarchy/structure → Org Chart / Dendrogram
│  ├─ Process stages → Funnel / Step-by-Step
│  ├─ Cyclical process → Cycle Diagram
│  └─ Strategy → SWOT Analysis
│
└─ SINGLE VALUE / KPI
   ├─ Just a number → Scaled-up Number
   ├─ Progress toward goal → Progress Bar / Waffle Chart
   └─ Value in range → Angular Gauge

---

Common Anti-Patterns

Don't	Do instead	Why
Pie chart with 8+ slices	Horizontal bar chart	Humans can't compare small angles
3D bar chart	2D bar chart	3D distorts perception, adds no data
Dual-axis chart	Two aligned charts or indexed lines	Two scales mislead about correlation
Rainbow color scale	Sequential or diverging palette	Rainbow has no perceptual order
Truncated y-axis bar chart	Start bars at zero	Non-zero baseline exaggerates differences
Scatter with 10,000+ points	Hexbin or 2D density	Overplotting hides the pattern
Line chart for unordered categories	Bar chart	Lines imply continuity between points
Area chart comparing series	Line chart or small multiples	Stacking occludes lower series

---

Response Format

When recommending a visualization, structure your response as:

### Recommended: [Chart Type]

**Why:** [1-2 sentences connecting the data shape to the chart's strengths]

**Alternative:** [Runner-up type] — better if [condition]

**Avoid:** [Common wrong choice] — because [reason]

**Implementation:**
- **D3.js**: [relevant module/example]
- **Observable Plot**: [relevant mark/transform]
- **Python (matplotlib/seaborn/plotly)**: [function]
- **R (ggplot2)**: [geom/function]
- **No-code**: [tool suggestion]

**Reference:** https://datavizproject.com/data-type/[slug]/

Adapt the implementation section to the user's known tech stack. If you know from context they use Python, lead with Python. If building for the web, lead with JS libraries.

---

Source

All visualization types and taxonomy from Data Viz Project by Ferdio, with supplementary guidance on selection and implementation.