如何将 pandas 数据帧转换为元组列表<string, list of></string,>

Question

I have a data frame which has a following structure

title      field1    field2   field3    field4   field5
title1    value11    value12  value13    value14   value15
title1    value21    value22  value23    value24   value25
title1    value31    value32  value33    value34   value35
title2    value1_1    value1_2  value1_3    value1_4   value1_5
title2    value2_1    value2_2  value2_3    value2_4   value2_5
title2    value3_1    value3_2  value3_3    value3_4   value3_5

I want to convert above data frame into a list of tuples<String, List of List>,

For example,

title1, [
    (value11, value12, value13, value14, value15),
    (value21, value22, value23, value24, value25),
    (value31, value32, value33, value34, value35)
]
title2, [
    (value1_1, value1_2, value1_3, value1_4, value1_5),
    (value2_1, value2_2, value2_3, value2_4, value2_5),
    (value3_1, value3_2, value3_3, value3_4, value3_5)
]

Answer 1

You can use tuple() inside lambda function in df.GroupBy.apply() as follows:

Assuming the fields columns are from the 2nd column onwards. If not, you can modify the column index range in df.columns[1:] below or explicitly use list the columns names below:

(df.groupby('title')[df.columns[1:]]
   .apply(lambda x: [tuple(y) for y in x.to_numpy()])
   .to_dict())

Result:

{'title1': [('value11', 'value12', 'value13', 'value14', 'value15'),
  ('value21', 'value22', 'value23', 'value24', 'value25'),
  ('value31', 'value32', 'value33', 'value34', 'value35')],
 'title2': [('value1_1', 'value1_2', 'value1_3', 'value1_4', 'value1_5'),
  ('value2_1', 'value2_2', 'value2_3', 'value2_4', 'value2_5'),
  ('value3_1', 'value3_2', 'value3_3', 'value3_4', 'value3_5')]}