如何通过前两个值对python列表/元组的第三个值进行分组/汇总/减少（不使用for循环）？

Question

What is the best way to group/aggregate/reduce python tuples in the following manner (without using nested for loops).

For example, given the following tuples:

entry1 = ('name1', 'surname1', 'product1')
entry2 = ('name1', 'surname1', 'product2')
entry3 = ('name1', 'surname1', 'product3')
entry4 = ('name2', 'surname2', 'product1')
entry5 = ('name2', 'surname2', 'product2')
entry6 = ('name2', 'surname2', 'product3')

How can the function func :

func(entry1, entry2, entry3, entry4, entry5, entry6)

return the data in a similar format to:

(('name1', 'surname1', ('product1', 'product2', 'product3')),
('name2', 'surname2', ('product1', 'product2', 'product3')))

Notes:

• The order of results in not important.
• The results can be a list or a tuple
• Enforcing product uniqueness is fine however not required. That is results such as ('name1', 'surname1', ('product1', 'product1', 'product2', 'product3')) are fine.

Answer 1

You can use collections.defaultdict to group first and later convert to a list or tuple:

from collections import defaultdict

def func(*args):
    d = defaultdict(list)
    for entry in args:
        d[entry[0], entry[1]].append(entry[2])
    return tuple((k[0], k[1], tuple(v)) for k, v in d.items())

Test:

entry1 = ('name1', 'surname1', 'product1')
entry2 = ('name1', 'surname1', 'product2')
entry3 = ('name1', 'surname1', 'product3')
entry4 = ('name2', 'surname2', 'product1')
entry5 = ('name2', 'surname2', 'product2')
entry6 = ('name2', 'surname2', 'product3')
print(func(entry1, entry2, entry3, entry4, entry5, entry6))

Result:

(('name1', 'surname1', ('product1', 'product2', 'product3')),
 ('name2', 'surname2', ('product1', 'product2', 'product3')))