8. 确定大学校园多样性

  1. #  US News给出的美国10所最具多样性的大学
  2.  In[71]: pd.read_csv('data/college_diversity.csv', index_col='School')
  3. Out[71]:

8. 确定大学校园多样性 - 图1

  1.  In[72]: college = pd.read_csv('data/college.csv', index_col='INSTNM')
  2.          college_ugds_ = college.filter(like='UGDS_')
  3.          college_ugds_.head()
  4. Out[72]:

8. 确定大学校园多样性 - 图2

  1.  In[73]: college_ugds_.isnull().sum(axis=1).sort_values(ascending=False).head()
  2. Out[73]: INSTNM
  3.          Excel Learning Center-San Antonio South         9
  4.          Philadelphia College of Osteopathic Medicine    9
  5.          Assemblies of God Theological Seminary          9
  6.          Episcopal Divinity School                       9
  7.          Phillips Graduate Institute                     9
  8.          dtype: int64
  1. #  如果所有列都是缺失值，则将其去除
  2.  In[74]: college_ugds_ = college_ugds_.dropna(how='all')
  3.  In[75]: college_ugds_.isnull().sum()
  4. Out[75]: UGDS_WHITE    0
  5.          UGDS_BLACK    0
  6.          UGDS_HISP     0
  7.          UGDS_ASIAN    0
  8.                        ..
  9.          UGDS_NHPI     0
  10.          UGDS_2MOR     0
  11.          UGDS_NRA      0
  12.          UGDS_UNKN     0
  13.          Length: 9, dtype: int64
  1. #  用大于或等于方法ge()，将DataFrame变为布尔值矩阵
  2.  In[76]: college_ugds_.ge(.15).head()
  3. Out[76]:

8. 确定大学校园多样性 - 图3

  1. #  对所有True值求和
  2.  In[77]: diversity_metric = college_ugds_.ge(.15).sum(axis='columns')
  3.          diversity_metric.head()
  4. Out[77]: INSTNM
  5.          Alabama A & M University               1
  6.          University of Alabama at Birmingham    2
  7.          Amridge University                     3
  8.          University of Alabama in Huntsville    1
  9.          Alabama State University               1
  10.          dtype: int64
  1. #  使用value_counts()，查看分布情况
  2.  In[78]: diversity_metric.value_counts()
  3. Out[78]: 1    3042
  4.          2    2884
  5.          3     876
  6.          4      63
  7.          0       7
  8.          5       2
  9.          dtype: int64
  1. #  查看哪些学校种群比例超过15%的数量多
  2.  In[79]: diversity_metric.sort_values(ascending=False).head()
  3. Out[79]: INSTNM
  4.          Regency Beauty Institute-Austin          5
  5.          Central Texas Beauty College-Temple      5
  6.          Sullivan and Cogliano Training Center    4
  7.          Ambria College of Nursing                4
  8.          Berkeley College-New York                4
  9.          dtype: int64
  1. #  用loc()方法查看对应行索引的行
  2.  In[80]: college_ugds_.loc[['Regency Beauty Institute-Austin', 
  3.                           'Central Texas Beauty College-Temple']]
  4. Out[80]:

8. 确定大学校园多样性 - 图4

  1. #  查看US News前五所最具多样性的大学在diversity_metric中的情况
  2.  In[81]: us_news_top = ['Rutgers University-Newark', 
  3.                'Andrews University', 
  4.                'Stanford University', 
  5.                'University of Houston',
  6.                'University of Nevada-Las Vegas']
  7.  In[82]: diversity_metric.loc[us_news_top]
  8. Out[82]: INSTNM
  9.          Rutgers University-Newark         4
  10.          Andrews University                3
  11.          Stanford University               3
  12.          University of Houston             3
  13.          University of Nevada-Las Vegas    3
  14.          dtype: int64

更多

  1. #  可以用最大种群比例查看哪些学校最不具有多样性
  2.  In[83]: college_ugds_.max(axis=1).sort_values(ascending=False).head(10)
  3. Out[83]: INSTNM
  4.          Dewey University-Manati                               1.0
  5.          Yeshiva and Kollel Harbotzas Torah                    1.0
  6.          Mr Leon's School of Hair Design-Lewiston              1.0
  7.          Dewey University-Bayamon                              1.0
  8.                                                                ... 
  9.          Monteclaro Escuela de Hoteleria y Artes Culinarias    1.0
  10.          Yeshiva Shaar Hatorah                                 1.0
  11.          Bais Medrash Elyon                                    1.0
  12.          Yeshiva of Nitra Rabbinical College                   1.0
  13.          Length: 10, dtype: float64
  1. #  查看Talmudical Seminary Oholei Torah哲学学校
  2.  In[84]: college_ugds_.loc['Talmudical Seminary Oholei Torah']
  3. Out[84]: UGDS_WHITE    1.0
  4.          UGDS_BLACK    0.0
  5.          UGDS_HISP     0.0
  6.          UGDS_ASIAN    0.0
  7.                        ... 
  8.          UGDS_NHPI     0.0
  9.          UGDS_2MOR     0.0
  10.          UGDS_NRA      0.0
  11.          UGDS_UNKN     0.0
  12.          Name: Talmudical Seminary Oholei Torah, Length: 9, dtype: float64
  1. #  查看是否有学校九个种族的比例都超过了1%
  2.  In[85]: (college_ugds_ > .01).all(axis=1).any()
  3. Out[85]: True