This Week in Toolkit Dev 2021/07/16

[JLockerman]

This Week in Toolkit Dev

Hi all, Toolkit dev lead here, this is the start of what we hope to be a regular series providing a look into what’s been happening recently in TimescaleDB Toolkit hyperfunctions development.
The biggest news for this week is the 1.0.0 release we did at the beginning of the week 🎉 . This marks the start of feature stabilization, precentile approximation—including percentile-agg, uddsketch, t-digest—and time-weighted average. With the big release behind us, we should return to our regular development schedule.

On to current dev:

• Our Hyperloglog++ implementation is in PR. This should improve the accuracy and space-usage of hyperloglog(). Once this is merged and tested, we believe that hyperloglog() will be ready for stabilization.
• We’ve started work on a pipeline API to help with the ergonomics of data analysis pipelines like

  timeseries |> downsample('5 minutes') |> asap(17)

• We’ve started design on some data-generation functions. These functions should help building tests and demos.

All in-progress issues are tracked here. Drop by below if you have any questions or comments!