Sorting a complex datastructure

Question

I have a list of tuples. Each tuple consists of another tuple and an index. I want to sort the list depending on the difference between the values of the inner tuple. Or, in another notation:

[((float, float), int), ...]

The two floats are what I want to use in the sort key.

I wrote this:

from math import fabs

def sort_for_avg(probabilities):  # 'probabilites' is a list of float-pairs
    return sorted(zip(probabilites, range(len(probabilites))),
                  key=lambda entry: fabs(entry[0][0]-entry[0][1]))

I tried working with itemgetter, but I didn't get the hang of it for something that isn't just sorting with a simple lookup. The 'sorted' list goes all over the place, seemingly randomly.

As the probability table can get quite big, I wanted my solution to not copy any values, which is why this is such a complicated one-liner. If someone knew how to do the same in a couple of lines with generators, I would be happy, too.

Could you provide some inputs and expected and actual outputs? Have a look at stackoverflow.com/help/mcve. It would be helpful if your "another notation" was Python's, e.g. [((float, float), int), ...] — jonrsharpe
– jonrsharpe, Commented Feb 2, 2015 at 16:36
Can not reproduce. After fixing the typos, this seems to work okay. Could the problem be somewhere else? Can you provide some test data? Is your result sorted randomly, or not sorted at all, i.e. is the index row still ordered? In this case, you probably forgot to assign the result back to the variable. — tobias_k
– tobias_k, Commented Feb 2, 2015 at 16:36

Imran · Accepted Answer · 2015-02-02 16:41:53Z

3

Example data:

probabilities = [((1.0,2.0),3),((3.0,1.0),1),((0.5,1.0),2)]

To sort in place:

probabilities.sort(key=lambda x: abs(x[0][0]-x[0][1]))

Output:

>>> probabilities
[((0.5, 1.0), 2), ((1.0, 2.0), 3), ((3.0, 1.0), 1)]

Or to return a new sorted list:

sorted_probabilities = sorted(probabilities, key=lambda x: abs(x[0][0]-x[0][1]))

Output:

>>> sorted_probabilities
[((0.5, 1.0), 2), ((1.0, 2.0), 3), ((3.0, 1.0), 1)]

edited Feb 2, 2015 at 16:41

answered Feb 2, 2015 at 16:36

Imran

13.6k8 gold badges69 silver badges82 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Arne Over a year ago

Thank you, this wraps it up nicely, everything working as expected now.

Collectives™ on Stack Overflow

Sorting a complex datastructure

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related