fatiando/verde

What it solves

Verde solves the problem of processing and interpolating spatial data—such as topography, point clouds, and geophysics surveys—onto a 2D surface (gridding). It provides a standardized, machine-learning-inspired interface for these operations, making it easier to integrate with the broader Python scientific stack.

How it works

Verde implements interpolation methods inspired by machine learning, using an interface similar to scikit-learn. It supports both Cartesian and geographic coordinates and can handle both scalar and vector data (such as GPS velocities or wind speed). The library also includes tools for data preparation and analysis, such as trend removal and blocked/windowed operations.

Who it’s for

Researchers and data scientists working with spatial data, topography, and geophysics who need to convert irregular point data into regular grids for analysis.

Highlights

• ML-inspired interface: Uses a scikit-learn-like API for gridding operations.
• Scientific stack integration: Fully integrates with numpy, pandas, scikit-learn, and xarray.
• Coordinate support: Works with both geographic and Cartesian coordinates.
• Analysis tools: Includes built-in support for cross-validation, trend removal, and blocked means.