Let's play together: Collaborative Data Science

on YouTube ↗24m

" summary

Companies often produce duplicated, fragile code instead of reusable libraries; R’s ecosystem has 13,000 CRAN packages but roughly single-author packages ~10,000, creating redundancy and newcomer friction. Excel import shows five competing packages and wrapper connectors; a concrete example merges two HDF5 implementations by combining high-level R code with a low-level C backend — merged HDF5 implementations — which results in measurable improvement: more downloads after merge. Recommendations: package code, use GitHub, enforce code review and dependency management.

AI-written summary · about

" chapters (5)

1. 0:00Start
2. 1:37Collaborative Data Science
3. 8:08Package Redundancy
4. 16:41Linus Torvalds
5. 19:20Breaking the Psychological Barrier

" related talks

• 04mThe Expanse - Navigating the R Package UniverseBen Arancibia · posit::conf(2024)
• 16mEXCELsioR!Cesaire Tobias · satRday Johannesburg 2019
• 15mAchieving a seamless workflow between R, Python and SAS from within RStudioMelissa Van Bussel · rstudio::conf 2019
• 21mDon’t repeat yourself, talk to yourself! Reporting with RSharla Gelfand · rstudio::conf 2019
• 17mOddly Satisfying - Find delight in the mundaneLiz Roten · rstudio::conf 2019
• 19mIt's a Great Time to be an R Package Developer!Hadley Wickham, Jenny Bryan · posit::conf(2023)
• 22mCultivating Your Own R Ecosystem as a Solo ContributorMeghan Hall · rstudio::conf 2019
• 15mDemystifying the art of creating custom libraries for your organizationDan Caley · rstudio::conf(2022)