smoking logo smoking: what about age

1 2 3 4 5 6 7
Notes

It seems that age indeed matters.

(clean_df
  .assign(r=1)
  .assign(age=lambda d: np.round(d['age'] / 5) * 5)
  .groupby(['age'])
  .agg(p_alive=('alive', np.mean),
        p_smokes=('smokes', np.mean))
  .plot());