8. 找到最常见的最大值

  1. #  读取college，过滤出只包含本科生种族比例信息的列
  2.  In[90]: pd.options.display.max_rows= 40
  3.  In[91]: college = pd.read_csv('data/college.csv', index_col='INSTNM')
  4.          college_ugds = college.filter(like='UGDS_')
  5.          college_ugds.head()
  6. Out[91]:

8. 找到最常见的最大值 - 图1

  1. #  用idxmax方法选出每行种族比例最高的列名
  2.  In[92]: highest_percentage_race = college_ugds.idxmax(axis='columns')
  3.          highest_percentage_race.head()
  4. Out[92]: INSTNM
  5.          Alabama A & M University               UGDS_BLACK
  6.          University of Alabama at Birmingham    UGDS_WHITE
  7.          Amridge University                     UGDS_BLACK
  8.          University of Alabama in Huntsville    UGDS_WHITE
  9.          Alabama State University               UGDS_BLACK
  10.          dtype: object
  1. #  用value_counts，查看最大值的分布
  2.  In[93]: highest_percentage_race.value_counts(normalize=True)
  3. Out[93]: UGDS_WHITE    0.670352
  4.          UGDS_BLACK    0.151586
  5.          UGDS_HISP     0.129473
  6.          UGDS_UNKN     0.023422
  7.          UGDS_ASIAN    0.012074
  8.          UGDS_AIAN     0.006110
  9.          UGDS_NRA      0.004073
  10.          UGDS_NHPI     0.001746
  11.          UGDS_2MOR     0.001164
  12.          dtype: float64

更多

  1. #  对于黑人比例最高的学校，排名第二的种族的分布情况
  2.  In[94]: college_black = college_ugds[highest_percentage_race == 'UGDS_BLACK']
  3.          college_black = college_black.drop('UGDS_BLACK', axis='columns')
  4.          college_black.idxmax(axis='columns').value_counts(normalize=True)
  5. Out[94]: UGDS_WHITE    0.670352
  6.          UGDS_BLACK    0.151586
  7.          UGDS_HISP     0.129473
  8.          UGDS_UNKN     0.023422
  9.          UGDS_ASIAN    0.012074
  10.          UGDS_AIAN     0.006110
  11.          UGDS_NRA      0.004073
  12.          UGDS_NHPI     0.001746
  13.          UGDS_2MOR     0.001164
  14.          dtype: float64