Reputation: 469
data= np.array([9,13,10,9,13,5,13,9,7,9,13,11])
fig9, ax9 = plt.subplots()
ax9.boxplot(data, notch=True)
I got the strange the notched boxplot. I want to use the notch to highlight the median. Why this figure showed the strange shape? Is this not strange?
I use MacOS 10.14.5 and python 3.7.2.
Upvotes: 5
Views: 606
Reputation: 844
It's explained quite clearly in the boxplot documentation. The note states:
Note: In cases where the values of the CI are less than the lower quartile or greater than the upper quartile, the notches will extend beyond the box, giving it a distinctive "flipped" appearance. This is expected behavior and consistent with other statistical visualization packages.
So if you calculate the params you will see that you are en the described case.
>>> [np.quantile(data, i) for i in np.arange(0.25, 1, 0.25)]
[9.0, 9.5, 13.0]
>>> scipy.stats as st
>>> st.t.interval(0.95, len(data)-1, loc=np.mean(data), scale=st.sem(data))
(8.425200299187184, 11.741466367479484)
Upvotes: 5