如何使用pandas从给定时间table创建每个主题的频率table？

Question

这是一个时间 table，列=小时，行=工作日，数据=主题 [工作日 x 小时]

                               1                      2                 3             4                 5                      6                      7
Name                                                                                                                                                   
Monday                   Project                Project           Project  Data Science  Embedded Systems            Data Mining  Industrial Psychology
Tuesday                  Project                Project           Project       Project      Data Science  Industrial Psychology       Embedded Systems
Wednesday           Data Science                Project           Project       Project           Project                Project                Project
Thursday             Data Mining  Industrial Psychology  Embedded Systems   Data Mining           Project                Project                Project
Friday     Industrial Psychology       Embedded Systems      Data Science   Data Mining           Project                Project                Project

如何生成 pandas.Dataframe 其中，行=工作日，列=主题，数据=相应工作日的主题频率？

必填table：[工作日 x 主题]

              Data Mining, Data Science, Embedded Systems, Industrial Psychology, Project                                                             
Name                                                                                                                                                   
Monday           1          1            1                 1                      3
Tuesday          ...         
Wednesday                     
Thursday                                     
Friday                               

        self.file = 'timetable.csv'
        self.sdf = pd.read_csv(self.file, header=0, index_col="Name")
        print(self.sdf.to_string())
        self.subject_frequency = self.sdf.apply(pd.value_counts)
        print(self.subject_frequency.to_string())
        self.subject_frequency["sum"] = self.subject_frequency.sum(axis=1)

Answer 1

使用 melt 展平您的数据框，然后 pivot_table 重塑您的数据框：

out = (
  df.melt(var_name='Freq', value_name='Data', ignore_index=False).assign(variable=1)
    .pivot_table('Freq', 'Name', 'Data', fill_value=0, aggfunc='count')
    .loc[df.index]  # sort by original index: Monday > Thuesday > ...
)

输出：

>>> out
Data       Data Mining  Data Science  Embedded Systems  Industrial Psychology  Project
Name                                                                                  
Monday               1             1                 1                      1        3
Tuesday              0             1                 1                      1        4
Wednesday            0             1                 0                      0        6
Thursday             2             0                 1                      1        3
Friday               1             1                 1                      1        3