2

The plot shown below, shows the percent of flights delayed by each Airline at an airport.

Violin Plot

From the above violin plot, all I could understand is $25^{th}$, $50^{th}$, $75^{th}$ percentiles are shown inside the leaf like structures(or body of a violin) shown with horizontal dotted lines. And presumably the peaks on top represent outliers(values more than 1.5*Inter Quartile Range).

Here are my doubts:

  1. What do the concentrations at the bottom mean? Can be seen for Alaska, Delta and United. And, what do the width of this concentration denote?
  2. What do multiple peaks represent? [This answer about violin plots][2] tells that violin plot is basically a histogram. So, all the bumps of the violin represent multiple modes that the data has(or peaks in the distribution)? (by bumps, I mean peaks and falls, for example, violin of American doesn't have any bumps in it, where as Alaska's violin has the most peaks and falls.)

This question weren't helpful for me too.

Naveen Reddy Marthala
  • 278

1 Answers1

2

  1. Vertical values are described by the y-axis title and tick labels: Daily % of Delayed Flights. So the masses Alaska, Delta, and JetBlue indicate that for some portion of the days measured, those three airlines had no delayed flights (i.e. had 0% of flights delayed). The width describes $2$ $\times$ the probability density, but also $2$ $\times$ the (identically distributed) frequency (neither of which have units or scales labeled on the x-axis in these plots, though they could).

  2. You are almost there: Even though its distribution of Daily % of Delayed Flights is a smooth curve, American does have a "bump" (or mode) at around 6.5% or 7%. So yes, the bumps represent peaks in the data and Alaska and Delta are the only two airlines with two peaks (Alaska's are at 0% and somewhere around 3.5%, and Delta's are at 0% and somewhere around 4%).

Alexis
  • 29,850