如何在Pandas DataFrame中对year列超2018的值执行减100的条件替换？

Hey there! Let's sort out that incorrect year value issue you're facing. Those 2067-style years (which should be 1967) are easy to fix with a few Pandas tricks. Here are three reliable methods to do the conditional replacement:

Method 1: Use loc for Explicit Indexing #

This is the most straightforward and Pandas-idiomatic way to target specific rows:

# Target rows where year > 2018, subtract 100 from the 'year' column
new_df.loc[new_df['year'] > 2018, 'year'] -= 100

The loc operator lets you precisely select the subset of data that needs updating—no risk of modifying unintended rows.

Method 2: Use numpy.where for Concise One-Liner #

If you prefer a more compact approach, numpy.where works perfectly here:

import numpy as np

# Syntax: np.where(condition, value_if_true, value_if_false)
new_df['year'] = np.where(new_df['year'] > 2018, new_df['year'] - 100, new_df['year'])

This replaces values that meet the condition with year - 100, and leaves all other values as-is in a single line.

Method 3: Use apply for Custom Logic (Good for More Complex Cases) #

While not as performant as the first two for large datasets, apply is flexible if you ever need to expand the logic later:

# Apply a lambda function to each value in the 'year' column
new_df['year'] = new_df['year'].apply(lambda x: x - 100 if x > 2018 else x)

The lambda function checks each year individually and adjusts it only if it's over 2018.

Verify the Fix #

After applying any of these methods, you should double-check that the issue is resolved:

# Check the updated statistics
print(new_df['year'].describe())

# Confirm no values are still above 2018
print(new_df[new_df['year'] > 2018])  # This should return an empty DataFrame

内容的提问来源于stack exchange，提问作者EdwardC