Function

[erikpc43]

Hello, i have a question about a function that I'm coding. I have a list that contains pairs like 

("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")

i would like to return a dictionary which keys would be the names of people and values should be everything the person "owns".

 

It should return a dictionary like this: 

{"Thomas": {"pen", "rubber"}, "Mike": {"pencilcase"}, "Tom": {"scissors"}}

 

What can i do here?

Edit: Im using python and i would like to write everyting in one line

[elpiop]

assuming you are using python, you could do something like this

 

from collections import defaultdict

d = defaultdict(set)
pairs = [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]
for name, item in pairs:
    d[name].add(item)
    
print(dict(d))

 

[erikpc43]

Yes this is one way to do it. But i would really like to write everything in one line.

[wasab]

9 minutes ago, erikpc43 said:

Yes this is one way to do it. But i would really like to write everything in one line.

Python compiler interpretor enforces indent so not possible. In java, everything can be squeeze into one line XD

[fizzlesticks]

2 minutes ago, wasab said:

Python compiler interpretor enforces indent so not possible. In java, everything can be squeeze into one line XD

Anything that can be written in Python can be written in 1 line, it might look silly sometimes though.

[wasab]

58 minutes ago, fizzlesticks said:

Anything that can be written in Python can be written in 1 line, it might look silly sometimes though.

What are you going to use for indent?

[0x21]

#!/usr/bin/env python3

from itertools import groupby

l = [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]

d = dict((k, [v[1] for v in vs]) for (k, vs) in groupby(sorted(l), lambda x: x[0]))

print(d)
# {'Mike': ['pencilcase'], 'Thomas': ['pen', 'rubber'], 'Tom': ['scissors']}

Or from modifying @elpiop's solution:

#!/usr/bin/env python3

from collections import defaultdict

l = [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]

d = defaultdict(set);[d[name].add(item) for (name, item) in l]

print(dict(d))
# {'Thomas': {'rubber', 'pen'}, 'Mike': {'pencilcase'}, 'Tom': {'scissors'}}

Edited November 25, 2018 by mshaugh
Added one-liner for @elpiop's solution

[fizzlesticks]

2 hours ago, wasab said:

What are you going to use for indent?

You use things that don't need indenting. For example here's elpiops answer in 1 line:

print(dict((lambda d: (d, [d[name].add(item) for name, item in [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]])[0])(__import__('collections').defaultdict(set))))

 

[wasab]

7 minutes ago, fizzlesticks said:

You use things that don't need indenting. For example here's elpiops answer in 1 line:

print(dict((lambda d: (d, [d[name].add(item) for name, item in [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]])[0])(__import__('collections').defaultdict(set))))

 

But what are you going to use to write a function? 

[fizzlesticks]

42 minutes ago, wasab said:

But what are you going to use to write a function? 

Lambdas mostly.

[reniat]

....why does this need to be 1 line? That's usually pretty bad codesmanship. Super complex one line statements might seem clever, but you get no performance benefit from it and it usually only makes things harder to maintain.

 

Good code is finding ways to be elegant and performant, all while also being easy to read, test, and maintain. Trying to find cheeky ways to crunch 4-5 lines into 1 PROBABLY isn't going to be "good" (though as always, there are exceptions)

Note, I don't write python professionally, so maybe there is some language standard i'm ignoring, but in a general software engineering sense this:

d = defaultdict(set)
pairs = [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]
for name, item in pairs:
    d[name].add(item)
    
print(dict(d))

is far better than:

print(dict((lambda d: (d, [d[name].add(item) for name, item in [("Thomas", "pen"), ("Mike", "pencilcase"), ("Thomas", "rubber"), ("Tom", "scissors")]])[0])(__import__('collections').defaultdict(set))))

And unless there is some python specific reason, I wouldn't let the 2nd example through code review. Sure it's not that hard to figure out what's going on, but why make the people maintaining this code have to go through even simple puzzles to figure out what your code is doing? What's the advantage of getting this all onto 1 line? The easier you make it for others to understand your code, the less chance a defect gets introduced. This is the same issue I have with list comprehensions that get too crazy, or when `auto` is abused in C++ to use a non-python example.