defaultdict python,根据键，使用defaultdict python进行聚合

I have a bunch of lines in text with names and teams in this format:

Team (year)|Surname1, Name1

e.g.

Yankees (1993)|Abbot, Jim

Yankees (1994)|Abbot, Jim

Yankees (1993)|Assenmacher, Paul

Yankees (2000)|Buddies, Mike

Yankees (2000)|Canseco, Jose

and so on for several years and several teams.

I would like to aggregate names of players according to team (year) combination deleting any duplicated names (it may happen that in the original database there is some redundant information). In the example, my output should be:

Yankees (1993)|Abbot, Jim|Assenmacher, Paul

Yankees (1994)|Abbot, Jim

Yankees (2000)|Buddies, Mike|Canseco, Jose

I've written this code so far:

file_in = open('filein.txt')

file_out = open('fileout.txt', 'w+')

from collections import defaultdict

teams = defaultdict(set)

for line in file_in:

items = [entry.strip() for entry in line.split('|') if entry]

team = items[0]

name = items[1]

teams[team].add(name)

I end up with a big dictionary made up by keys (the name of the team and the year) and sets of values. But I don't know exactly how to go on to aggregate things up.

I would also be able to compare my final sets of values (e.g. how many players have Yankee's team of 1993 and 1994 in common?). How can I do this?

Any help is appreciated

解决方案

You can use a tuple as a key here, for eg. ('Yankees', '1994'):

from collections import defaultdict

dic = defaultdict(list)

with open('abc') as f:

for line in f:

key,val = line.split('|')

keys = tuple(x.strip('()') for x in key.split())

vals = [x.strip() for x in val.split(', ')]

dic[keys].append(vals)

print dic

for k,v in dic.iteritems():

print "{}({})|{}".format(k[0],k[1],"|".join([", ".join(x) for x in v]))

Output:

defaultdict(,

{('Yankees', '1994'): [['Abbot', 'Jim']],

('Yankees', '2000'): [['Buddies', 'Mike'], ['Canseco', 'Jose']],

('Yankees', '1993'): [['Abbot', 'Jim'], ['Assenmacher', 'Paul']]})

Yankees(1994)|Abbot, Jim

Yankees(2000)|Buddies, Mike|Canseco, Jose

Yankees(1993)|Abbot, Jim|Assenmacher, Paul