Announcement

Collapse
No announcement yet.

MOWOG 6 Run Statistics

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • MOWOG 6 Run Statistics

    So I wanted to learn how to use the DataFrames package in Julia, so I decided to put it to use going through MOWOG 6 Results.

    The photo album reduces the resolution, so I uploaded them an an album here https://photos.app.goo.gl/M9irTqK3Q2A6taMj2
    Last edited by BergeGuy; 10-11-2017, 04:14 PM.
    Daniel Berge #8
    2005 Subaru WRX STi

  • #2
    Daniel Berge #8
    2005 Subaru WRX STi

    Comment


    • #3
      Originally posted by BergeGuy View Post
      So I wanted to learn how to use the DataFrames package in Julia, so I decided to put it to use going through MOWOG 6 Results.
      What is the "Y" axis?
      John Mensch

      2011 Driver of the Year.
      Proud recipient of the 2008 Paul Weidner "Spirit of the Sport" award.
      Founder of F.rugal R.acing E.nterprises. (FRE) "It's not F.R.E.E., but it's close....."

      Quote section;
      Resentment: It's like taking poison and waiting for the other person to die...
      "If there is no plan, what can go wrong!" - Some Clown
      "I like to think of myself as thrifty, cheap sounds so.......cheap." - Same Clown
      To quote a wise man I once knew; "If you can't have a little fun what good are you?"

      Comment


      • #4
        Originally posted by the tick View Post

        What is the "Y" axis?
        I'm gonna say they're histograms and that's the fraction of people in the corresponding time segment.

        Very cool work up Dan. Definitely highlights the learning taking place.
        Jay McKoskey
        1991 Miata #12/63

        Comment


        • #5
          Yeah, it's basically a normal distribution, with some magic that I don't actually fully understand the math, that works with the time scale. So essentially, the area under any curve is effectively 1.0.
          Daniel Berge #8
          2005 Subaru WRX STi

          Comment


          • #6
            this is what im going to school for quality and chart data (industrial Engineering) ... very cool to actually see it applied to a passion ...

            Comment


            • #7
              This is a learning experience for me, trying to do something practical with it. So reading into this more, the results are are Kernel Density Estimation from the data. So the results should be taken with skepticism. So in general, the I believe the KDE is giving the confidence of a time being correct for the entire set. So for the morning runs for a veteran autorcrosser, the probability of a 50.0s run was the most probable, but by run #3, a veteran autocrosser was even more likely to run a 52 second run, with an overall narrower distribution. This may seem counter-intuitive, but it's really pointing to the drivers who were upper 50s getting their times down to the mid 50s for a tigher grouping.
              Daniel Berge #8
              2005 Subaru WRX STi

              Comment


              • #8
                To that point, can you use it to chart lap time improvement across runs by driver?

                Comment


                • #9


                  That is a good question. This Violin plot is interesting, but it get's quickly cluttered if you try and track the improvement between more than two groups.
                  Daniel Berge #8
                  2005 Subaru WRX STi

                  Comment

                  Working...
                  X