Understand representation
Identify which question the data can answer.
AURORA LAB
Market data experiments.
A public learning collection on representation, event order and reproducible reconstruction. Read a note, inspect an experiment and run the source.
Open collection / 01COLLECTION 01
From events to a
reconstructed state.
For researchers and engineers working with event data. Basic programming knowledge is helpful; the examples use synthetic data and require no market-data subscription.
Python uses only the standard library. Browser experiments need no account.
Make event order explicit
Compare source and receive order in a working replay.
Inspect queue position
See how aggregate depth hides an order-level distinction.
Record the experiment
Document inputs, invariants and information boundaries.
FROM EXPERIMENT TO RESEARCH
Reproducible by design
The experiments make data assumptions visible: what a feed records, how an event changes state and what can be checked on replay. Source files are included so that the result can be inspected beyond the browser.
For work with production data, agree the data rights, sequence semantics and validation criteria before extending a demonstration into a research pipeline.
Discuss Aurora Lab