如何用Pandas保留行仅清除指定列数据（CSV数据清洗）

Got it, let's adjust your script to keep the rows intact while only clearing the firstname, lastname, and email fields for matching customers. Here's a clean, efficient way to do it:

Step 1: Read the Opt-Out Customer List #

First, make sure you've loaded the list of hashed customer IDs that need cleaning:

import pandas as pd

# Read the opt-out customer CSV and extract the hashed_customer column as a list
cust_opt_out_id = pd.read_csv('path/to/your/opt_out_customers.csv')['hashed_customer'].tolist()

Step 2: Target and Clear Specific Columns #

Instead of filtering out rows, use Pandas' .loc indexer to directly update the relevant fields for matching customers. This is way more efficient than looping with iterrows() for large datasets:

# Locate rows where hashed_customer is in the opt-out list, then set specified columns to None
df_in.loc[df_in['hashed_customer'].isin(cust_opt_out_id), ['firstname', 'lastname', 'email']] = None

Step 3: Export the Cleaned Data #

When saving back to CSV, use na_rep='NULL' to ensure empty values show up as NULL (matching your desired output):

# Export the cleaned DataFrame to CSV
df_in.to_csv('cleaned_customer_orders.csv', index=False, na_rep='NULL')

Why This Works #

• The .loc indexer lets you target a subset of rows (matching opt-out customers) and a subset of columns (the personal info fields) in one go.
• Setting values to None ensures Pandas recognizes them as missing values, which we can format as NULL during export.
• This avoids the inefficiency of looping through every row with iterrows(), which is slow for large datasets.

What Was Wrong With the Original Script #

Your original code df_cust_out = df_in[~df_in['hashed_eater_uuid'].isin(cust_opt_out_id)] was filtering out entire rows that matched the opt-out list. We don't want to remove the rows—just erase the personal data while keeping the rest of the order details.

内容的提问来源于stack exchange，提问作者JMV12