10. 创建、删除列

  1. #  通过[列名]添加新列
  2.  In[96]: movie = pd.read_csv('data/movie.csv')
  3.  In[97]: movie['has_seen'] = 0
  4.  In[98]: movie.columns
  5. Out[98]: Index(['color', 'director_name', 'num_critic_for_reviews', 'duration',
  6.                 'director_facebook_likes', 'actor_3_facebook_likes', 'actor_2_name',
  7.                 'actor_1_facebook_likes', 'gross', 'genres', 'actor_1_name',
  8.                 'movie_title', 'num_voted_users', 'cast_total_facebook_likes',
  9.                 'actor_3_name', 'facenumber_in_poster', 'plot_keywords',
  10.                 'movie_imdb_link', 'num_user_for_reviews', 'language', 'country',
  11.                 'content_rating', 'budget', 'title_year', 'actor_2_facebook_likes',
  12.                 'imdb_score', 'aspect_ratio', 'movie_facebook_likes', 'has_seen'],
  13.                dtype='object')
  1. #  给新列赋值
  2.  In[99]: movie['actor_director_facebook_likes'] = (movie['actor_1_facebook_likes'] + 
  3.                                               movie['actor_2_facebook_likes'] + 
  4.                                               movie['actor_3_facebook_likes'] + 
  5.                                               movie['director_facebook_likes'])
  6.  In[100]: movie['actor_director_facebook_likes'].isnull().sum()
  7. Out[100]: 122
  1. #  用all()检查是否所有的布尔值都为True
  2.  In[101]: movie['actor_director_facebook_likes'] = movie['actor_director_facebook_likes'].fillna(0)
  3.  In[102]: movie['is_cast_likes_more'] = (movie['cast_total_facebook_likes'] >= 
  4.                                   movie['actor_director_facebook_likes'])
  5.  In[103]: movie['is_cast_likes_more'].all()
  6. Out[103]: False
  8.  In[104]: movie = movie.drop('actor_director_facebook_likes', axis='columns')
  9.  In[105]: movie['actor_total_facebook_likes'] = (movie['actor_1_facebook_likes'] + 
  10.                                        movie['actor_2_facebook_likes'] + 
  11.                                        movie['actor_3_facebook_likes'])
  13.           movie['actor_total_facebook_likes'] = movie['actor_total_facebook_likes'].fillna(0)
  14.  In[106]: movie['is_cast_likes_more'] = movie['cast_total_facebook_likes'] >= \
  15.                                   movie['actor_total_facebook_likes']
  17.           movie['is_cast_likes_more'].all()
  18. Out[106]: True
  1.  In[107]: movie['pct_actor_cast_like'] = (movie['actor_total_facebook_likes'] / 
  2.                                 movie['cast_total_facebook_likes'])
  3.  In[108]: movie['pct_actor_cast_like'].min(), movie['pct_actor_cast_like'].max() 
  4. Out[108]: (0.0, 1.0)
  6.  In[109]: movie.set_index('movie_title')['pct_actor_cast_like'].head()
  7. Out[109]: movie_title
  8.           Avatar                                        0.577369
  9.           Pirates of the Caribbean: At World's End      0.951396
  10.           Spectre                                       0.987521
  11.           The Dark Knight Rises                         0.683783
  12.           Star Wars: Episode VII - The Force Awakens    0.000000
  13.           Name: pct_actor_cast_like, dtype: float64

更多

  1. #  用insert()方法原地插入列
  2.  In[110]: profit_index = movie.columns.get_loc('gross') + 1
  3.           profit_index
  4.  In[111]: movie.insert(loc=profit_index,
  5.                  column='profit',
  6.                  value=movie['gross'] - movie['budget'])
  7.  In[112]: movie.head()
  8. Out[112]:

10. 创建、删除列 - 图1