Interactive workbenches for gaining familiarity with viz libs

Back in Feb 2018 I’d had enough of my Swiss Cheese brain trying & failing to retain any familiarity with the parameters available in key Python viz libraries:

  • MatPlotLib
  • Seaborn
  • others fetched up from the deep when neither of the above was able to deliver an intuitive experience.

I’d also encountered Zeppelin notebooks for the first time…

… and, since these come pre-packaged with some fairly impressive viz capabilities out of the box, felt any further time spent on Python in this realm would be wasted. I decided to capture some favourite configurations in a repository for my future self and to share with others:

During build-out of the notebook in this repository I bumped into a friend from GA who reminded me we’d all been keen on interactive widgets we’d seen presented at PyLondinium and thought we might come up with ways of jazzing up our notebooks with those:

See what you make of it! github.com/olisteadman/intviz

Next steps

  • [DONE] Grab everything off AzureNotebooks where it was initially hosted (the service has now been cancelled)
  • Find out how to run a SageMaker instance with password protection so it can still be accessed publicly but not run ad hoc

DataCat @AstraZeneca. Bass/BV @StornowayBand. Fan @lowislandmusic. CTO @Tigmuso. Voc/Gtr @DrachmaBand. DataEng @ICISOfficial. DataSys @GBioinf. DataSci @GA.