SUMS: some useful models
  • Merge asof

    2026-01-01

    Merge asof fuzzy matches two timeseries. Typically you have X = (tx,x1,x2,) and Y = (ty, y1, y2,) data matrices and the timestamp vectors tx, ty are not aligned. For each time in tx, there exists a value in ty that is an approximate match. Often we want a data matrix Z = (tx, ty, x1-n, y1-n) so that we can compute joint statistics on (X,Y) where Z = merge_asof(X,Y). …

    data-science

  • First post

    2026-01-01

    Each short specific post is about a useful technique found in stats, ML, trading, data science, or computing. I skip the intro, focus on the kernel, and assume you can fill in the blanks. The focus is on tools that I’ve actually used and return to regularly in my work or hobbies.

SUMS: some useful models

Powered by Hugo & Notepadium