Skip to main content

Visualizations

Turn your data into charts, graphs, and dashboards—all created through natural language.

What are Visualizations?​

Visualizations are charts and graphs powered by your Views and Sources. Instead of wrestling with charting tools, you describe what you want in plain English, and the AI creates it. From simple bar charts to complex Sankey diagrams, all charts are built on the Vega visualization grammar for unlimited flexibility.

Visualization Example

A chart created from a simple natural language request

Why Visualizations Matter​

Instant insights: See patterns and trends that are invisible in tables.

Publication-ready: Export high-resolution charts for presentations and reports.

Interactive: Charts update automatically when underlying data changes.

Unlimited flexibility: If a chart type exists, Shadowfax can create it.

No design skills needed: Describe your vision, the AI handles the details.

How to Create Visualizations​

Natural Language Creation​

The easiest way:

@[monthly_revenue] Create a line chart showing revenue trend over time

The AI will:

  1. Understand which data to use
  2. Choose an appropriate chart type
  3. Configure axes and styling
  4. Generate the visualization in 15-45 seconds

Chart Generation

AI generates charts from your description

Using the /visualize Command​

For more control, use the /visualize slash command:

/visualize @[sales_by_region] as a horizontal bar chart with regions
sorted by total sales

This mode gives the AI explicit instructions about chart structure, sorting, and styling.

Creating from Any Data Object​

You can visualize:

  • Sources: Original data files or database tables
  • Views: Transformed datasets
  • Input Objects: Variable tables or scenario data

Right-click any node and select "Visualize" for quick chart creation.

Chart Types Available​

Shadowfax supports any Vega-compatible chart, including:

Basic Charts

  • Bar charts (vertical, horizontal, stacked, grouped)
  • Line charts
  • Area charts
  • Scatter plots
  • Pie and donut charts

Statistical Charts

  • Histograms
  • Box plots
  • Violin plots
  • Distribution curves

Advanced Visualizations

  • Heatmaps
  • Treemaps
  • Sunburst diagrams
  • Sankey diagrams (flow visualization)
  • Network diagrams
  • Geographic/choropleth maps
  • Bubble charts
  • Waterfall charts

Chart Type Gallery

Gallery of available chart types

Customizing Charts​

Conversational Editing​

Modify charts naturally:

Make the bars blue
Rotate the x-axis labels 45 degrees
Add data labels to each bar
Increase font size to 14px
Add a subtitle showing the date range

The AI understands context—you don't need to re-describe the entire chart.

Multi-edit with Lists​

Request multiple changes at once:

Update the chart:
- Change color scheme to green gradient
- Add grid lines
- Remove the legend
- Make the title bold

Draft Mode for Edits​

Chart modifications create a draft version. You'll see:

  • Current chart
  • Proposed changes
  • "Update chart" button to apply

This lets you preview before committing.

Chart Draft Mode

Preview chart changes before applying

Managing Visualizations​

Location in Workbook​

Charts appear:

  • Below View nodes: As chart icons beneath the node that created them
  • In the Visualizations section: Left sidebar listing all charts
  • In the bottom panel: Toggle to "Viz" tab to see full-size

Multiple Charts per View​

One View can power multiple charts:

  • Monthly revenue as a line chart
  • Same data as a bar chart
  • Same data as a data table with sparklines

Multiple Charts

One View powering multiple visualization styles

Quick Access​

Right-click for options:

  • Edit visualization
  • Export (PNG, SVG, JSON)
  • Duplicate chart
  • Delete chart

Use Cmd+K: Quick AI editing without full chat

Exporting Visualizations​

PNG Export: High-resolution images for presentations SVG Export: Vector format for print and design tools JSON Export: Raw Vega spec for developers

All exports maintain quality and formatting.

Export Options

Export dialog with format options

Common Use Cases​

Trend analysis: Line charts showing metrics over time

Comparison: Bar charts comparing categories or groups

Distribution: Histograms showing data spread

Composition: Pie charts for part-to-whole relationships

Correlation: Scatter plots showing variable relationships

Flow analysis: Sankey diagrams for multi-step processes

Geographic patterns: Maps showing regional data

Complex relationships: Network diagrams for connections

Tips & Best Practices​

Start simple: Request a basic chart first, then customize. Easier than describing everything upfront.

Be specific about sorting: If order matters, say "sorted by revenue descending"

Request meaningful colors: "Color by category" or "Use a red-to-green gradient for performance"

Think about labels: Ask for "data labels on bars" or "percentage labels" if needed

Consider your audience: Mention if it's for a presentation (bigger fonts) or dashboard (compact)

Iterate quickly: Small adjustments are faster than recreating from scratch

Export early: Grab PNG exports as you go—easy to regenerate if needed

Understanding Chart Processing Time​

Chart generation takes 15-45 seconds depending on:

  • Data size
  • Chart complexity
  • Number of customizations requested

You'll see a progress indicator while the AI works.

Processing Indicator

Progress indicator during chart generation