Python删除DataFrame多列异常：仅删除一列的解决建议咨询

Hey there! Let's dig into why only one column got deleted instead of the ~800 you targeted. Here are the most common culprits and fixes to get this sorted:

1. You're not modifying the DataFrame correctly in loops #

If you wrote code like this to delete columns one by one:

for col in cols_to_drop:
    df.drop(col, axis=1)

This won't work! By default, drop() returns a new copy of the DataFrame instead of modifying the original. You have two options to fix this:

• Assign the result back to your DataFrame every time:

  for col in cols_to_drop:
      df = df.drop(col, axis=1)
  

• Use the inplace=True parameter (note: this modifies the DataFrame directly and is less recommended in newer pandas versions):

  for col in cols_to_drop:
      df.drop(col, axis=1, inplace=True)
  

That said, looping to delete 800 columns is inefficient—skip to point 4 for a better approach.

2. Most columns in your cols_to_drop list don't exist in the DataFrame #

It’s possible that only one column name in your list actually matches a column in your DataFrame. To verify this, check how many of your target columns are present:

existing_cols = [col for col in cols_to_drop if col in df.columns]
print(f"Number of columns that exist in the DataFrame: {len(existing_cols)}")

If this number is 1, that’s exactly why only one column got deleted. You’ll need to fix your cols_to_drop list to ensure the column names match exactly (case-sensitive, no typos!).

3. You mixed up the axis parameter #

Double-check that you’re using axis=1 in your drop() call. axis=0 targets rows, not columns. If you accidentally used axis=0, you might have deleted a row instead—but if you’re seeing a column missing, this is less likely, but worth ruling out.

4. Delete all target columns in one go (recommended!) #

The best way to delete multiple columns is to pass the entire list to drop() in a single call. This is way faster and avoids loop-related mistakes:

# Delete columns, ignoring any that don't exist (avoids errors)
df = df.drop(cols_to_drop, axis=1, errors='ignore')

# If you want to know which columns weren't deleted (no ignore)
try:
    df = df.drop(cols_to_drop, axis=1)
except KeyError as e:
    print(f"Columns not found: {e}")

Start by verifying your cols_to_drop list matches the DataFrame’s column names, then use the single-call method—it’ll save you time and headaches!

内容的提问来源于stack exchange，提问作者user15051990