Data Visualization

Data Visualization — DBX Studio

Chart Types Available

The generate_chart tool supports these types:

Workflow

1. Understand what the user wants to visualize
2. Write the SQL query to get the data (data_query)
3. Call generate_chart with the config
4. Confirm chart title and axes are meaningful

generate_chart Parameters

{
  "chart_type": "bar",
  "title": "Monthly Revenue by Product Category",
  "x_axis": "category",
  "y_axis": "revenue",
  "data_query": "SELECT category, SUM(amount) AS revenue FROM orders GROUP BY 1 ORDER BY 2 DESC",
  "group_by": "category"
}

Chart Selection Guide

User says "trend" or "over time" → line chart, x_axis = date column User says "compare" or "by category" → bar chart User says "breakdown" or "share" → pie chart (only if ≤ 7 categories) User says "distribution" or "spread" → histogram User says "relationship" or "correlation" → scatter

Data Query Patterns

Bar: Top N categories

SELECT category, COUNT(*) AS count
FROM orders
GROUP BY category
ORDER BY count DESC
LIMIT 10

Line: Time series

SELECT DATE_TRUNC('day', created_at) AS date, SUM(amount) AS revenue
FROM orders
WHERE created_at >= NOW() - INTERVAL '30 days'
GROUP BY 1
ORDER BY 1

Pie: Proportion breakdown

SELECT status, COUNT(*) AS count
FROM orders
GROUP BY status

Design Principles

• Always give the chart a descriptive title including the time period if relevant
• Keep x_axis and y_axis names human-readable (not raw column names)
• For large result sets, aggregate before charting (avoid raw row-level data)
• Pie charts: max 7 slices, group remainder as "Other"

Source: Dbxstudio/dbx-studio — distributed by TomeVault.