What’s Coming in Lathe 0.1.0?

Lathe’s Easy To Master Machine-Learning Module for Julia is getting an Upgrade!

Lathe is currently in version 0.0.8 LTS and 0.0.9 Unstable. But with this versioning there are already plans laid out for Lathe up until 0.1.1, and as Lathe 0.0.9 screeches closer and closer to release, it’s very exciting to see where Lathe started, and what is going to become of Lathe. For those who aren’t in the loop, Lathe is a machine-learning module for Julia that makes it easy and productive to deploy real-time solutions for data examination and machine-learning deployment.

Lathe has several benefits to traditionally used packages, like Sklearn. In addition to supporting machine-learning algorithms, Lathe also has a lot of data control tools inside of Lathe.preprocess. These tools include Scalars, Encoders, TestTrain Split, SortSplit, and many more. Another advantage is the built-in statistics module, Lathe.stats. Lathe.stats includes an array of statistical tests, some distributions, standard — inferential — and Bayesian statistics, as well as rank-sum tests and correlations. In addition, Lathe now has on board validation, which was moved from Lathe.validate to Lathe.stats. Something to note is that Lathe.validate will be deprecated as part of our package flow, which is new and we’ll get into later.