@alfcnz We used a similar closure strategy in dataflow notebooks where only expressions and assignments in the last line are promoted to globals: https://t.co/KXABFv2fzZ Also see @minrk's closure cell magic: https://t.co/Ba04bisx0d
@scheidegger@domoritz@ijlyttle Using code from modules/packages without clawing my eyes out? ES modules seem nice but then we have old stuff with require and stuff that only runs in node, etc.
@scheidegger Agree about the concerns of whatโs going on underneath and docs. However, having switched to it for about a year now, I like a lot of the benefits. Used a lot of emacs, but canโt claim to have been an expert
@sharoz As someone who hasnโt done much with #rstats, I found it amusing that in a thread on readable code, my initial thought was to wonder why data attributes were being encrypted ๐
@scheidegger I still have the original site open in a tab, and nyt apparently changed today's word. Guessed nyt answer (wrong) and had to determine which more obscure word was left
@_hex_tech Great to see these ideas gaining acceptance and more polish. We championed dataflow notebooks as a solution in the talk right after Joel's JupyterCon talk: https://t.co/HJ0gsAabyT
My #JupyterCon2020 talk on structuring notebooks around outputs goes live at 12pm EDT. See how ipycollections improves output display and dfnotebook-extension enables dataflow execution in #JupyterLab https://t.co/HJ0gsAabyT
Nice solution to syntax + efficiency: "At first glance table expressions look like normal JavaScript functions... but hold on! Arquero takes a set of function definitions, maps them to strings, then parses, rewrites, and compiles them to efficiently manage data internally."
@duto_guerra@arnicas@ireneros@vlandham I translated the original learnjsdata to @observablehq a while ago. Hope to revisit this soon for D3 6 and new libraries like @DanfoJs., but free to fork and update for ES6: https://t.co/dghGYHru2D
@nathancarter5 Thanks! I'm happy to report that I'm spending the summer at @argonne_lcf (remotely), getting back to work on these ideas with the scientists there. Stay tuned for some work with #jupyterlab and other extensions of the ideas.