In [ ]:
import hvplot.pandas  # noqa

box plots are most useful when grouped by additional dimensions.

In [ ]:
from bokeh.sampledata.sprint import sprint as df

df.head()
In [ ]:
boxplot = df.hvplot.box(y='Time', by='Medal', height=400, width=400, legend=False)
boxplot

Overlay this plot with the jittered scatter plot of the medalist times using the * operator:

In [ ]:
boxplot * df.hvplot.scatter(y='Time', x='Medal', c='orange').opts(jitter=0.5)

Use groupby to create a separate plot for each medal type with a widget for selecting between the plots.

In [ ]:
df.hvplot.box(y='Time', groupby='Medal', by='Country', ylabel='Sprint Time', height=400, width=600)