Building on our robust text-to-SQL API, version 1.13.0 introduces powerful visualization features, enabling users to not only generate and run SQL queries but also to visualize the results dynamically within conversational applications. This integration significantly enhances user interaction with data, facilitating a more intuitive understanding and presentation of insights.
- Interactive Visualization: Users can interact with the generated visualizations through natural language commands. This allows for dynamic manipulation of visual data, such as:
- Zooming In/Out: Focus on specific data points or zoom out for a broader view.
- Changing Visualization Styles and Types: Switch between different types of charts (e.g., from a bar chart to a scatter plot) or change the visual style of charts to match presentation needs.
- Applying Filters: Narrow down data points interactively to focus on relevant subsets of the data.
- Supported Chart Types:
- Bar Chart: Ideal for comparing quantities across different categories.
- Line Chart: Perfect for displaying trends over time.
- Pie Chart: Useful for showing proportions within a whole.
- Scatter Plot: Effective for identifying data patterns and correlations.
- Area Chart: Highlights the magnitude of trends over time by filling the area under the line.
- Histogram: Useful for showing frequency distributions.
- Heatmap: Ideal for visualizing data density or intensity.
- Bubble Chart: Useful for comparing three dimensions of data.
- Visualization Formats: Charts can be rendered in multiple formats to suit various application requirements:
- Vega-Lite: For lightweight, declarative visualizations.
- Pyplot: Utilizes Python's matplotlib for detailed and customizable visual representations.
- Images: For straightforward, static visual outputs.
- Custom Formats: Allows flexibility for integrating unique styling and branding requirements.
Example Usage:
- Natural Language Interaction: A user could say, "Show me a bar chart of sales by product category, then convert it to a pie chart," and the system would execute the appropriate SQL query, generate a bar chart, and then transform it into a pie chart upon request.
This release marks a significant advancement in the capability of conversational tools to provide not just textual or tabular responses but rich, interactive visual insights. This is ideal for rapid decision-making and enhances the overall user experience by making data interaction more engaging and accessible.