This repository hosts notes/notebooks and one off explorations, so I can look them up easily.
The posts are in reverse chronological order by the date they were added, though most date back to further before.
- date:
2022-06-27
- topic: Jupyter notebook that addresses some of the SQL syntax quirks mentioned in the against-sql post by Jamie Brandon. Uses the
FunSQL
python library. - tiny-notes: Using a DSL hosted in the language allows for richer abstractions. I really need to clean up the FunSQL library code though.
- date:
2022-06-02
- topic: Implements the VMSP (Vertical Maximal Sequence Patterns) algorithm to find frequent subsequences in a collection of event logs. Based off the work here.
- tiny-notes: Could be used to surface frequent patterns when digging into event sequences. The MAQUI work for recursive event exploration uses it. Tabled for later when I try to replicate the full application.
- date:
2020-04-20
- topic: What I can understand from the pitch for
Metric layer
tools. - tiny-notes: Seems like a nice way to keep analytics disciplined. Not sure if it warrants a separate service though, maybe as a library that you can integrate with your existing analytics code.
- date:
2021-12-20
- topic: Notes on what I like about node-and-wire editors for working with data.
- tiny-notes: What I would really like is the Jupyter workflow, but cells backed to a content addressable storage. So, very much like observable. In fact, don't stop at data science, make even regular programming work that way. Ex - Unison.
- date:
2021-11-01
- topic: Describes the recommendation system behind an adaptive learning API platform for school students.
- tiny-notes: The post never made it to the company blog, so stored here for posterity.
- date:
2021-10-29
- topic: What writing the Snake game in your calculator is like; following the tutorial at link.
- tiny-notes: Concatenative programming seems pretty cool. Time to dive into a language like APL.
- date:
2021-05-30
- topic: How materialize does change data capture; original post on their blog.
- tiny-notes: pretty neat model, was easy enough to follow and write it up in python.