将Pandas DataFrame的genre列拆分为多个布尔列的实现方法

Great question! Working with pipe-separated categorical columns like your genre field is a common data wrangling task, and pandas has some handy tools to make this easy. Let's tackle both of your problems step by step.

1. Get All Unique Genre Names #

First, we need to extract every distinct genre from the pipe-separated strings. Here's how to do it efficiently with pandas:

First, load your CSV into a DataFrame:

import pandas as pd

# Replace 'movies.csv' with your actual file path
df = pd.read_csv('movies.csv')

Next, split each genre entry into a list, "explode" those lists into individual rows, then grab the unique values:

# Split genre strings into lists, then explode to create one genre per row
all_genres = df['genre'].str.split('|').explode()

# Get unique genres, sorted for readability
unique_genres = sorted(all_genres.unique())

# Print the result to check
print("Unique genres:", unique_genres)

This will give you a clean, sorted list of every genre present in your dataset. If there are missing values in the genre column, add df['genre'] = df['genre'].fillna('') first to avoid errors.

2. Create Boolean Columns for Each Genre #

Once you have the unique genres, generating boolean (True/False) columns is a breeze. There are two main approaches depending on your needs:

Approach 1: Use str.get_dummies() (Quickest) #

Pandas' built-in str.get_dummies() method automatically creates columns for each unique genre, with 1 indicating the genre is present and 0 otherwise. You can convert these to boolean values if you prefer:

# Create dummy columns from pipe-separated genres
genre_dummies = df['genre'].str.get_dummies(sep='|')

# Optional: Convert 0/1 to True/False
genre_dummies = genre_dummies.astype(bool)

# Merge the new columns with your original DataFrame
df_with_booleans = pd.concat([df, genre_dummies], axis=1)

This will add columns named after each genre (e.g., Adventure, Science Fiction) with boolean values.

Approach 2: Manual Column Creation (Customizable) #

If you want more control over column names (like using is_adventure instead of Adventure), loop through the unique genres we found earlier:

# First, make sure we have our unique_genres list from step 1
for genre in unique_genres:
    # Create a column like 'is_adventure' with True if the genre is present
    df_with_booleans[f'is_{genre.lower().replace(" ", "_")}'] = df['genre'].str.contains(genre, case=False)

This creates lowercase, snake_case column names and uses case-insensitive matching (so "adventure" and "Adventure" are treated the same). Adjust case=True if you need exact case matching.

Both methods work efficiently even with 10,000 rows—pandas is optimized for these kinds of operations.

内容的提问来源于stack exchange，提问作者EdwardC