A box plot draft that deleted possible outliers

This AI draft compares the weights (kg) of parcels that arrived at the dorm mailroom (Dorm) and the office mailbox (Office) with box plots. The draft drew the box plots from kept, a DataFrame with the parcels heavier than the upper bound removed.

Do not delete possible outliers before you check the original records. Change one line so that the box plots are drawn with all the parcels.

Code to fix

import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns

desks = ["Dorm"] * 8 + ["Office"] * 8
kg = [2, 3, 3, 4, 4, 5, 6, 21, 1, 2, 2, 3, 5, 6, 7, 9]
parcels = pd.DataFrame({"desk": desks, "kg": kg})

q1 = parcels["kg"].quantile(0.25)
q3 = parcels["kg"].quantile(0.75)
upper = q3 + 1.5 * (q3 - q1)
kept = parcels[parcels["kg"] <= upper]  # remove outliers

fig, ax = plt.subplots()
sns.boxplot(data=kept, x="desk", y="kg", ax=ax)
ax.set_ylabel("Weight (kg)")
plt.show()