Kotlin/dataframe

What it solves

Kotlin DataFrame addresses the tension between Kotlin's strict static typing and the dynamic, often unpredictable nature of structured data (like JSON or CSVs). It allows developers to process data with the flexibility of a dynamic library while maintaining type safety and null safety.

How it works

The library provides a functional, immutable data processing pipeline where operations are chained together using a natural-language-like DSL. It utilizes a compiler plugin to generate extension properties on the fly, enabling type-safe access to data columns. It is designed to work both in standard JVM projects and interactively within Jupyter notebooks or REPLs, where it can handle hierarchical data structures and integrate with the Kotlin Notebook environment.

Who it’s for

It is designed for JVM developers and data scientists using Kotlin who need to perform structured data manipulation, cleaning, and analysis, particularly those looking for a pandas-like experience within the Kotlin ecosystem.

Highlights

• Type-Safe Access: Generates extension properties for safe data access and null safety.
• Hierarchical Data Support: Capable of representing complex structures like JSON or trees of JVM objects.
• Interoperable: Easily converts to and from Kotlin data classes, collections, and formats like Apache Arrow, JSON, and CSV.
• Functional API: Uses an immutable approach where every transformation returns a new DataFrame instance.
• Notebook Integration: Deeply integrated with Kotlin Notebook and Jupyter for interactive data exploration.