underdarkGIS’s avatarunderdarkGIS’s Twitter Archive—№ 15,263

    1. I'm really glad that GI journals have started to require a data&code package. However, so far, I haven't been able to completely reproduce any of the result sections of the papers I've been asked to review. #reproducibility #giscience 🧵
      Permalink On twitter.com Twitter logo underdarkGIS’s avatar ♻️ 20 Retweets ❤️ 112 Favorites Mood +3 🙂
  1. …in reply to @underdarkGIS
    We are in this weird transitional period, where most papers describe work that was started without reproducibility in mind. The general mindset seems to be "runs on my machine, so should be ok ;-P"
    On twitter.com Twitter logo underdarkGIS’s avatar ♻️ 2 Retweets ❤️ 13 Favorites Mood -2 🙁
    1. …in reply to @underdarkGIS
      It is clear from the errors I've encountered in reproduction that none of the authors have bothered with trying to reproduce their own results on a second machine.
      Permalink On twitter.com Twitter logo underdarkGIS’s avatar ♻️ 1 Retweets ❤️ 7 Favorites Mood -3 🙁
      1. …in reply to @underdarkGIS
        It's also sad when there is code but it requires a proprietary software package that I don't have access to. It would be good to know that in advance and maybe send the paper on to someone else who can actually run the code.
        Permalink On twitter.com Twitter logo underdarkGIS’s avatar ♻️ 1 Retweets ❤️ 28 Favorites Mood +1 🙂