Upstream libraries Tablesaw and Smile have been updated! One piece of big news here really is that ElasticNet, which is a great regression model and used all over the place is now part of the main Smile codebase. There are new releases of: * tech.datatype - https://github.com/techascent/tech.datatype * tech.ml.dataset - https://github.com/techascent/tech.ml.dataset * http://tech.ml - https://github.com/techascent/tech.ml Finally, I wrote a lengthy piece of documentation on exactly how the sequence/reader concept developed in the tech.datatype library applies to doing arithmetic operations with both a basic performance analysis and a real world example from client work: * https://github.com/techascent/tech.datatype/blob/master/docs/01-operations.md Using this library one thing that is interesting is you can do elementwise vector operations without leaving clojure datastructures:

user> (require '[tech.v2.datatype :as dtype])
nil
user> (require '[tech.v2.datatype.functional :as dfn])
nil
user> (dfn/+ 4 4)
8
user> (dfn/+ 4 (list 1 2 3 4 5))
(5 6 7 8 9)
user> (instance? tech.v2.datatype.ObjectReader *1)
false
user> (instance? Iterable *2)
true
user> (dfn/+ 4 [1 2 3 4 5])
[5 6 7 8 9]
user> (instance? tech.v2.datatype.ObjectReader *1)
true
user> (instance? Iterable *2)
true