Sorting a Python Dictionary by Priority

Question

I have a python dictionary consisting of (names,value) pairs like so

pyDictionary = {"Bob":12,"Mellissa":12,"roger":13}

What i would like to do is to obtain a sorted version of the above dictionary where the sorting is done by giving first prority to the value and if the value for two pairs are the same then the comparision should be made by lexographically comparing the names.

How can i acheive this in python3.7?

Don't use `dict` as a variable name, since doing so overrides the standard definition. — Tom Karzes, May 16 '19 at 19:59
Notice that order in python dictionaries is random. If you want to sort dictionary use `OrderedDict` from collections — Relandom, May 16 '19 at 19:59
@Relandom The OP specifically states that the question is for Python 3.7, where the order of dict keys are preserved. — blhsing, May 16 '19 at 20:01
@AtifFarooq No need to apologize - it's just a Python tip :) — Tom Karzes, May 16 '19 at 20:05

yatu · Accepted Answer · 2019-05-16T20:09:37.323

7

You can use sorted with a key, and build an OrderedDict from the result to mantain order.

(the last step is only necessary with python 3.6 <, in Python 3.7 dicts are ordered on their key insertion time)

from collections import OrderedDict
d = {"Mellissa":12, "roger":13, "Bob":12}

OrderedDict(sorted(d.items(), key=lambda x: (x[1], x[0])))
# dict(sorted(d.items(), key=lambda x: (x[1], x[0]))) # for Python 3.7
# [('Bob', 12), ('Mellissa', 12), ('roger', 13)]

Or you can also use operator.itemgetter to directly fetch the value and key from each tuple respectively:

OrderedDict(sorted(d.items(), key=itemgetter(1,0)))
# dict(sorted(d.items(), key=itemgetter(1,0))) # python 3.7
# [('Bob', 12), ('Mellissa', 12), ('roger', 13)]

edited May 16 '19 at 20:09

answered May 16 '19 at 19:58

yatu

86,083
12
84
139

In *Python 3.7* *dict*s are ordered on their key insertion time. – CristiFati May 16 '19 at 20:04
Yes @CristiFati added a comment in the solution – yatu May 16 '19 at 20:06
What if i want to sort the first column in Descending but the second in ascending order can this solution be extended to this new scenario? – atifcppprogrammer May 16 '19 at 20:31
You mean the key from a to z and values descending @AtifFarooq ? – yatu May 16 '19 at 20:37
Use `dict(sorted(d.items(), key=lambda x: (-x[1], x[0])))` – yatu May 16 '19 at 20:39
@yatu Thanks i figured it out on my own but thanks for your reponse ! – atifcppprogrammer May 16 '19 at 21:16

score 2 · Answer 2 · answered May 16 '19 at 19:59

2

You can sort the dict items with a key function that reverses the order of the key-value tuple:

dict(sorted(pyDictionary.items(), key=lambda t: t[::-1]))

answered May 16 '19 at 19:59

blhsing

91,368
6
71
106

What about [**`reversed`**](https://docs.python.org/3/library/functions.html#reversed)? – Peter Wood May 16 '19 at 20:06
1

`reversed` won't work because it returns an iterable, which is not comparable. – blhsing May 16 '19 at 20:08

Sorting a Python Dictionary by Priority

2 Answers2