Axle A Spire based DSL
summary
Axle is a Scala-embedded DSL built on Spire that provides concrete type classes and small shims so algorithms actually run: Axle core depends only on Spire and Discipline, axle-jblas supplies linear algebra witnesses so regressions and matrix-based algorithms are playable, and a units library represents quantities with unit-aware arithmetic. Edit distance is implemented as a metric space and passes triangle-inequality tests. A Spark Monte Carlo π example runs locally using a custom bunker aggregator but encounters serialization edges; documentation via tut remains a priority.
AI-written summary · about
chapters (18)
- 0:00Intro
- 0:46Learning the category theory
- 1:55Mono 8s
- 2:41Algebra
- 3:43Abstract Algebra
- 4:23Project Goals
- 7:41Documentation
- 8:18Other things
- 8:49Test suite
- 9:54Unit of measurement
- 14:25Edit distance
- 15:02Distance function
- 16:34Spark demo
- 17:13Spark RDD
- 18:26Unico Sigma
- 18:52Spark Stuff
- 19:19Serialization
- 21:36Visualization
related talks
36mData pipelines engineering made simple with Scala
42mScala: The Unpredicted Lingua Franca for Data Science
26mA Functional EDSL for Economics Concepts & Models
27mEmpowering PySpark with Lance Format for Multi-Modal AI Data Pipelines
59mSpark 2.0
32mRecursion Schemes Without the Barbed Wire (Part 3)
03mDBTB INT Andrew Or r
09mBackstage interview