Example 1: Basic 3D Scatter with Fitted Surfaces

This example shows 3D scatter points with automatically fitted surfaces using kernel smoothing.

Features:

• Three groups: Each group has different underlying patterns (saddle, paraboloid, plane)
• Default smoothing: Each group uses its optimal smoothing parameter
• Interactive controls: Toggle surfaces/points, adjust X/Y ranges, select smoothing levels
• Color coding: Each group has a distinct color for both points and surface

Example 2: Custom Surface Fitting Function

This example demonstrates using a custom surface fitting function.

The custom fitter uses a simple moving average approach with adjustable bandwidth.

• Custom smoother: User-defined function for fitting surfaces
• Multiple smoothing levels: Pre-compute surfaces at different smoothing parameters
• Slider control: Switch between smoothing levels or use group-specific defaults

Example 3: Financial Data with Multiple Grouping Schemes

This example demonstrates stock returns analysis with switchable grouping schemes.

Dataset: 150 synthetic stocks with financial metrics

• X-axis (Carry): Interest rate differential - higher carry suggests positive yield
• Y-axis (Trading Activity): Recent market volatility and volume
• Z-axis (24h Return): Expected stock return over next 24 hours (%)

Grouping Schemes: Use the dropdown to switch between different groupings:

• Industry: Group by GICS sector (Technology, Financials, Healthcare, Energy, Consumer, Industrials)
• Country: Group by geographic region (USA, UK, Germany, Japan, China)

Key Insights:

• Each grouping reveals different patterns in the data
• Technology and Healthcare sectors tend to show higher returns
• Different countries have different return profiles based on local market conditions
• The L1/L2 toggle allows switching between robust (median) and smooth (mean) fits

Summary

ScatterSurface3D combines 3D scatter plots with fitted surfaces:

Key Features

• Fitted Surfaces: Automatically fit smooth surfaces to point clouds
• Multiple Smoothing Levels: Pre-compute surfaces at different smoothing parameters
• Group-Specific Defaults: Each group can have its own optimal smoothing parameter
• Interactive Controls:
  • Toggle all surfaces on/off
  • Toggle all points on/off
  • Toggle individual groups by clicking color buttons
  • Adjust X and Y ranges with sliders
  • Select smoothing parameter or use group defaults
• Customizable: Provide your own surface fitting function

Use Cases

• Visualizing complex 3D relationships in data
• Comparing fitted surfaces across different groups
• Exploring optimal smoothing parameters interactively
• Non-parametric regression visualization

Controls Guide:

• Smoothing dropdown: Select "Defaults" to use group-specific smoothing, or choose a specific value to apply to all groups
• Color buttons: Click to toggle visibility of that group's points and surface
• X/Y Range: Filter data to focus on specific regions