Winsorize and Cap Outliers with df.clip() in Pandas

Question

Given a `delivery_time_min` column with extreme outliers (e.g. 5,000 minutes due to sensor errors), winsorize the data by capping values at the 99th percentile upper bound and 1st percentile lower bound.

Accepted Answer

Optimal Solution:
import pandas as pd

lower_bound = df['time_min'].quantile(0.01)
upper_bound = df['time_min'].quantile(0.99)

# Winsorize / Cap between bounds
df['capped_time_min'] = df['time_min'].clip(lower=lower_bound, upper=upper_bound).round(2)

result_df = df[['capped_time_min']]

Technical Explanation:
`.clip(lower, upper)` trims values outside the thresholds to the exact boundary limits, maintaining sample size without dropping rows.

time_min
12
15
18
20
22
25
28
30
35
40
5000

capped_time_min
12.3
15
18
20
22
25
28
30
35
40
4504