New LensKit release

[Michael Ekstrand]

I just tagged & pushed 0.7.

Major changes:

• RecListAnalysis has the ability to make sure all users, even those without recommendations, have entries in the results
• Parallelism and performance improvements. User-user is much faster; item-item training is a little slower but requires significantly less memory.
• Some lesser-used arguments, particularly for the batch routines, are now keyword-only.
• We use JobLib now for parallelizing batch evaluation.

Upcoming things to look out for:

• a recommend-and-measure API to compute recommendation metrics when it isn't feasible to produce all recommendation lists in advance (e.g. full-length recommendation lists 
• an demo experiment

Enjoy!

- Michael

-- 

Michael D. Ekstrand — michael...@boisestate.edu — https://md.ekstrandom.net
Assistant Professor, Dept. of Computer Science, Boise State University

People and Information Research Team (PIReT) — http://coen.boisestate.edu/piret/

I may send mail outside of working hours; I do not expect you to.