What follows is an overview of my experiments that i have published into a GitHib repo. The “Examples” folder are what i would term “simple learnings” and “Full Scripts” are scripts that to a lesser or greater extent do something “useful”. Im also not suggesting that anything here is “best practice” or method A performs better than method B, I simply do not have the required size of data to make that call. My aim was to learn the language.
TLDR: Check out the script MovieLens09-CosineSimilarityFromCSVWithMax.usql for a U-SQL movie recommender.
U-SQL was introduced last year, but word of mouth about the language has been quite limited to date. I’ll be interested in seeing what other examples pop up over the next few months.