How to sort list of dictionaries by value

Question

dictionary[pattern_key] = {"key": index_key, "document": index_source, "startPos":index_start, "endPos": index_end}

This is an extract of my list of dictionaries

{
'AGACAATCTC': {'startPos': '174', 'document': 'source-document01012.txt', 'endPos': '183', 'key': 'AGACAATCTC'}, 
'GGTCAGACAA': {'startPos': '18', 'document': 'source-document01012.txt', 'endPos': '27', 'key': 'GGTCAGACAA'}, 
'TAGATGAAGT': {'startPos': '102', 'document': 'source-document01012.txt', 'endPos': '111', 'key': 'TAGATGAAGT'}
}

How can i sort that by document and then by startPos ?

i tried something like this but does not work

sorted_dict = sorted(dictionary, key=itemgetter(pattern_key[document]))

script.py

#!/usr/bin/env python
import sys

dictionary = {};

for pattern in sys.stdin:

    if "," in pattern:
        pattern_key, pattern_source, pattern_start, pattern_end = pattern.strip().split(",")
        index_file =  open('index.txt', 'r')

        for line in index_file:
            if "," in line:
                index_key, index_source, index_start, index_end = line.strip().split(",")
                if pattern_key == index_key:
                    dictionary[pattern_key] = {"document": index_source, "startPos":index_start, "endPos": index_end}

sorted(dictionary.items(), key = lambda x: (x[1]['document'], int(x[1]['startPos'])))

for k, v in dictionary.items():
    print (k, '-->', v)

For output, do you want the entire dict values, or just the keynames? ['AGACAATCTC', 'TAGATGAAGT'] like that? — akash12300
– akash12300, Commented Mar 27, 2016 at 16:27
i need the entire output sorted by document and then startPos... — Petros Kyriakou
– Petros Kyriakou, Commented Mar 27, 2016 at 16:28
Sorted does NOT update the dictionary to have sorted values. You need to use the list of (<key>, <dict-val>) tuple returned by sorted() command. If you convert list of tuples to dictionary and then use it, it may again disturb the sorted order. — akash12300
– akash12300, Commented Mar 27, 2016 at 16:56
I don't see a list of dict. but instead a dict. of dict. or am I missing something here? — Iron Fist
– Iron Fist, Commented Mar 27, 2016 at 17:14
If you want a dictionary to be sorted, you'll need to use collections.OrderedDict. Regular dictionaries do not guarantee the order of the keys. — Jared Goguen
– Jared Goguen, Commented Mar 27, 2016 at 17:17

Stewart Smith · Accepted Answer · 2016-03-27 18:52:09Z

3

You can get the entries in the inner dictionary as keys for sorted:

sorted(dictionary.items(), key = lambda x: (x[1]['document'], int(x[1]['startPos'])))

A tuple key will be sorted first by the 0th element, then 1st, and so on.

Note that this produces a list of tuples, where each tuple is (str, dict).

EDIT:
In your context, the correct implementation is the following:

sorted_values = sorted(dictionary.items(), key = lambda x: (x[1]['document'], int(x[1]['startPos'])))

for k, v in sorted_values:
    print (k, '-->', v)

edited Mar 27, 2016 at 18:52

answered Mar 27, 2016 at 16:15

Stewart Smith

1,41615 silver badges32 bronze badges

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

6 Comments

Stewart Smith Over a year ago

Running this on your sample data returns:

('GGTCAGACAA', {'document': 'source-document01012.txt', 'endPos': '27', 'startPos': '18', 'key': 'GGTCAGACAA'}) ('TAGATGAAGT', {'document': 'source-document01012.txt', 'endPos': '111', 'startPos': '102', 'key': 'TAGATGAAGT'}) ('AGACAATCTC', {'document': 'source-document01012.txt', 'endPos': '183', 'startPos': '174', 'key': 'AGACAATCTC'})

. This is sorted by document and startPos.

akash12300 Over a year ago

Can you give an example, for what value it did not sort using 'startPos'? Because I tried similar approach and it worked for me.

Petros Kyriakou Over a year ago

hmm weird, it gives me a randomized output?

('AAAGCTTACA', '-->', {'startPos': '132', 'document': 'source-document01012.txt', 'endPos': '141'}) ('GGAGAAATCT', '-->', {'startPos': '78', 'document': 'source-document01012.txt', 'endPos': '87'}) ('TCGGGAGCAA', '-->', {'startPos': '216', 'document': 'source-document01012.txt', 'endPos': '225'}) ('CGGTTTATGT', '-->', {'startPos': '204', 'document': 'source-document01012.txt', 'endPos': '213'}) ('TCACGTAGGA', '-->', {'startPos': '234', 'document': 'source-document01012.txt', 'endPos': '243'})

i removed "key" just for clarity, do not mind the "-->"

akash12300 Over a year ago

No the sort works. Are you sure everything else is in place and proper?

Petros Kyriakou Over a year ago

will post my whole script

|

Iron Fist · Accepted Answer · 2016-03-27 17:36:50Z

Make your sorting based on your desired criteria then create a new OrderedDict from the sorted list, since dict cannot keep the sorting by it's nature:

>>> from collections import OrderedDict
>>>
>>> d = {'AGACAATCTC': {'endPos': '183', 'document': 'source-document01010.txt', 'key': 'AGACAATCTC', 'startPos': '174'}, 'GGTCAGACAA': {'endPos': '27', 'document': 'source-document01010.txt', 'key': 'GGTCAGACAA', 'startPos': '18'}, 'TAGATGAAGT': {'endPos': '111', 'document': 'source-document01011.txt', 'key': 'TAGATGAAGT', 'startPos': '102'}}
>>> 
>>> d_ordered = OrderedDict(sorted(d.items(), key=lambda t:(t[1]['document'], int(t[1]['startPos']))))
>>> 
>>> d_ordered
OrderedDict([('GGTCAGACAA', {'endPos': '27', 'document': 'source-document01010.txt', 'key': 'GGTCAGACAA', 'startPos': '18'}), ('AGACAATCTC', {'endPos': '183', 'document': 'source-document01010.txt', 'key': 'AGACAATCTC', 'startPos': '174'}), ('TAGATGAAGT', {'endPos': '111', 'document': 'source-document01011.txt', 'key': 'TAGATGAAGT', 'startPos': '102'})])

Collectives™ on Stack Overflow

How to sort list of dictionaries by value

2 Answers 2

6 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

6 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related