#tsfeatures — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #tsfeatures, aggregated by home.social.
-
Marker: Too much computer. Needs more information in marker's domain of knowledge.
Me, adding features: I have univariate time series feature analysis working across multiple dimensions, and a mostly functional classification system attached. Each dimension runs in its own thread, shit is fast yo. I have attained enlightenment.
Anyone looking for eldrich R knowledge for time series analysis, theft is an interesting CRAN package. Blunt instruments are great. Note tsibbles have header data, you can't just yank columns out of them, it causes baffling downstream errors even if keys and indexes are included.