search

JohannesWalter / BostonAdventures

Testing before using Git with John
0 stars 0 forks source link

Outliers, outliers #1

Open JanBiermann opened 1 year ago

JanBiermann commented 1 year ago

(Note: This thing mainly exists because I wanted to create an issues ;)).

I was trying to better understand why outliers seems to play such a role in figure 6 & 7 (the ones I have described) and in figure 3 (the one you have described) and how to reconcile theses differences in my head. Maybe we should talk about this at some point.

One thing, I have noticed: Graph 3 limits the x-axis.

Without limiting the axis, it looks more in line with what I see in figures 6 & 7, i.e. there are quite some guesses which are way too high. This looks like long tail when looking at the density and has a stong influence in the other graph, when looking at the average (as a point estimate).

grafik

Anyway, the bigger point of this text here is: Assigning issues is fun! See you around, ninja turtle!

EDIT: after reading your part. I can see how outliers don't matter. Your part seems to have more insight than mine. But even if we decide to kick out my part all together: I guess we will run into aspect, where outliers will play a big role (simply looking at averages)

JohannesWalter commented 1 year ago

Note: This comment mainly exists to reply to an issue.

We have to ask John about how to deal with your graphs concerning outliers.