Documentation
Build typed table pipelines.
Start with the essentials, then use the topics below as a compact map of the language and runtime.
Guides by topic
Getting Started
Install Ibex, use the REPL, and build a first table pipeline.
Language guide
Table clauses, joins, reshaping, aggregation, windows, types, and nulls.
Language walkthrough
Detailed examples for expressions, joins, nulls, reshaping, streaming, and models.
I/O guide
CSV, JSON, Parquet, SQLite/ADBC, Kafka, and schema hints.
Notebook integration
Use Ibex from Python, R, and native application workflows.
Notebook examples
Detailed Python and R notebook sessions, bindings, and persistent state.
Advanced workflows
Compilation, streams, producer queues, and model specifications.
Performance and editor support
Benchmark context, scaling details, and VS Code setup.
Reference
A concise index of language functionality and the specification.
Installation
Download a release or build Ibex from source. The installation guide includes both paths.
Quickstart
Read a table, transform it, and print a result in the first pipeline guide.
REPL
Use the REPL to inspect tables, run expressions, and load scripts. See the REPL walkthrough.
Language
Pipelines
Tables flow through bracketed clauses such as filter, select, and update. Use the language guide for an overview and the language walkthrough for detailed examples.
Expressions and types
Expressions combine columns, literals, operators, and calls. Ibex has static scalar, series, table, and time-frame types.
Functions
Use built-ins for common data work and typed fn declarations for reusable code. Browse the function reference.
Control flow
Use expressions, case, and typed functions to define transformations; pipelines keep data work declarative.
Data
Reading & writing
CSV, JSON, Parquet, and other data sources are provided through plugins. Read the I/O guide.
Joins, reshaping, aggregation, and windows
The language guide covers joins, binding, reshaping, grouped operations, and TimeFrames.
Integration
Python
Use Ibex from notebook workflows while keeping tables and scalar bindings typed. See the notebook integration guide and detailed notebook examples.
R
The R package offers a lazy dplyr backend with an explicit fallback boundary. See the R integration guide.
C++ / native compilation
Compile an Ibex program to C++23 for native deployment, or connect a custom source through an extern function. See the integration guide.
Streams and advanced workflows
Continuous pipelines, producer queues, native compilation, and model specifications are covered in the advanced workflows guide.
Tooling
CLI
Run scripts or start the interactive REPL with the ibex executable.
Configuration
Configure plugin paths and runtime options when starting Ibex.