During the last meetup a few interesting ideas/discussions came up like interactive visualizations of Open Government Data, Geocoding APIs, derstandard.at datasets and Hadoop vs Spark. During the second part of the meetup Mateuz held a great talk covering R-package creation, Github and CI.


Even without my full memory capacity I’ll try to sum them up by project ;):

  • Open Government Data Vienna
    • With the nice visualizations from Christian in mind (Population Densities and Bike Parking Lots) it would be nice to make up a list of features to be plotted from https://www.data.gv.at and make this visualization interactive (e.g. using Shiny).
    • Mateuz also introduced us to the topic of Geocoding and its numerous APIs (e.g. from Google) and suggested to write a package for it.
  • Text Mining Vienna
    • After the discussion about puclic newspapers in Vienna Bernhard already provided a small dataset to play around with. We’ll get our hands dirty and provide ideas (topic modelling, mood index, popular words, etc.) next time!
    • Bernhard & Roland mentioned interest to talk about Hadoop -> Spark -> SparkR. How to set things up, integrate R, etc. Sounds like a great topic next years first meetup.
    • Hillary Clintons E-mails: Some ideas are there - focus will be on the newspapers project for now.
  • Fantasy Sports, looks abandoned by Daniel?


Mateusz covered all important topics about modern R package creation including the following points:

  • Setting up Github repo for startup package.
  • Create R package skeleton
  • Include test cases
  • Include continous integration
  • Include code coverage

Materials are available at http://zozlak.org/Setting_up_a_first_R_project.html.


See you next time for a punchR!