如何用Replace函数将DataFrame中的XNA值替换为NaN？

No problem at all! Since pandas doesn’t natively recognize 'XNA' as a missing value, using the replace() method is the straightforward solution here. Here's how to implement it for your large DataFrame:

Basic Full DataFrame Replacement #

First, make sure you have numpy imported (since np.nan is pandas' standard missing value marker):

import pandas as pd
import numpy as np

# Replace all 'XNA' entries with NaN
df.replace('XNA', np.nan, inplace=True)

• The inplace=True parameter modifies your original DataFrame directly. If you’d rather create a copy instead of altering the original, omit inplace=True and assign the result to a new variable:

  cleaned_df = df.replace('XNA', np.nan)
  

Replace in Specific Columns Only #

If you don’t want to replace 'XNA' across all 36 columns (e.g., only in certain columns where it logically represents a missing value), you can target specific columns:

# List of columns where you want to replace 'XNA'
target_cols = ['column_name_1', 'column_name_2']
df[target_cols] = df[target_cols].replace('XNA', np.nan)

Handling Case Variations (Optional) #

If your data has case variations like 'xna' or 'Xna', use a regex to catch all instances:

import re

# Case-sensitive exact match
df.replace(r'^XNA$', np.nan, regex=True, inplace=True)

# Case-insensitive match
df.replace(r'^xna$', np.nan, regex=True, flags=re.IGNORECASE, inplace=True)

Once you run this, df.isnull().sum() will correctly count all former 'XNA' values as nulls. This operation is efficient enough for your 1.6M-row DataFrame—pandas handles these vectorized operations quickly even with large datasets.

内容的提问来源于stack exchange，提问作者Nikunj Patel