Question
In the context of this exercise, why are there two symmetric halves for the violin plot? Do we have to count both?
Answer
Each half of the violin plot shows the same thing. We do not have to count both toward the data distributions.
The reason for the halves is essentially just for visual effect, which leads to the name since it is shaped like a violin.
As a result, removing one half will still let us read the data distribution. However, we must keep in mind the values in the axis, since each half is mirrored and will be in the reverse order of the other side.