CODEWITHSUNDEEP
djangodjango-templates
YardenST
YardenST

Reputation: 5257

django regroup and than dictsort by list length

I need to regroup a list, than to show the results ordered by the number of items per result's list.

something like this:

{%  regroup car_models by make as car_models_by_make %}

{% for make in car_models_by_make|dictsort:"???" --> the problem %}
    {{ make.grouper }}  ({{ make.list|length }}) <br>
{% endfor %}

so if the data is:

[
('panda','fiat'),
('500','fiat'),
('focus','ford')
]

the result should be:

fiat 2
ford 1

I tried: dictsort:"make.list.count","make.list.length","make.list|length" .. none did the work..

Upvotes: 1

Views: 2375

Answers (3)

YardenST
YardenST

Reputation: 5257

Using this template tags I've made:

@register.filter()
def groups_sort(groups):

    groups.sort(key=lambda group: len(group['list']))
    return groups

@register.filter()
def groups_sort_reversed(groups):
    groups.sort(key=lambda group: len(group['list']), reverse=True)
    return groups

It is possible to do this:

{% load custom_templatetags %}
{%  regroup car_models by make as car_models_by_make %}

{% for make in car_models_by_make|groups_sort_reversed %}
    {{ make.grouper }}  ({{ make.list|length }}) <br>
{% endfor %}

will give you

fiat 2
ford 1

Using groups_sort will give you:

ford 1
fiat 2

this of course works on Django Models too

Enjoy

Upvotes: 1

K Z
K Z

Reputation: 30453

You should pass your data to the template as a sorted list of dicts:

>>> data = [ ('panda','fiat'),
... ('500','fiat'),
... ('focus','ford') ]
>>> 
>>> from itertools import groupby

# note that we need to sort data before using groupby
>>> groups = groupby(sorted(data, key=lambda x: x[1]), key=lambda x:x[1])

# make the generators into lists
>>> group_list = [(make, list(record)) for make, record in groups]

# sort the group by the number of car records it has
>>> sorted_group_list = sorted(group_list, key=lambda x: len(x[1]), reverse=True)

# build the final dict to send to the template
>>> sorted_car_model = [{'make': make, "model": r[0]} for (make, records) in sorted_group_list for r in records]
>>> sorted_car_model
[{'make': 'fiat', 'model': 'panda'}, {'make': 'fiat', 'model': '500'}, {'make': 'ford', 'model': 'focus'}]

The goal here is to sort the list such that the car makers are ranked by the numbers of models they produce.

then, use:

{% regroup car_models by make as make_list %}

{% for item in make_list %}
    {{ item.grouper }} ({{ item.list|length }}) <br>
{% endfor %}

to get:

fiat 2
ford 1

Referece: regroup

If your data looks like the one you have in your comments: 

>>> car_models = [ 
...     {'make': 'ford', 'model': 'focus'},
...     {'make': 'fiat', 'model': 'panda'},
...     {'make': 'ford', 'model': '500'},
...     {'make': 'fiat', 'model': '500'},
...     {'make': 'opel', 'model': '500'},
...     {'make': 'opel', 'model': '500'},
...     {'make': 'opel', 'model': '500'},
... ] 
>>> 
>>> groups = groupby(sorted(car_models, key=lambda x:x['make']), key=lambda x:x['make'])
>>> groups = [(make, list(x)) for make, x in groups]
>>> sorted_group_list = sorted(groups, key=lambda x:len(x[1]), reverse=True)
>>> sorted_car_model = [{'make': make, 'model': r['model']} for (make, records) in sorted_group_list for r in records]

>>> sorted_car_model
[{'make': 'opel', 'model': '500'}, {'make': 'opel', 'model': '500'}, {'make': 'opel', 'model': '500'}, {'make': 'fiat', 'model': 'panda'}, {'make': 'fiat', 'model': '500'}, {'make': 'ford', 'model': 'fo
cus'}, {'make': 'ford', 'model': '500'}]

Alternatively, you can just pass the results to the template like follows:

>>> from collections import Counter
>>> l = [r['make'] for r in car_models]
>>> c = Counter(l)
>>> result = [{k: v} for k,v in dict(c).iteritems()]
>>> result = sorted(result, key=lambda x: x.values(), reverse=True)
[{'opel': 3}, {'fiat': 2}, {'ford': 2}]

Then, pass this result to the template and simply just render the results instead:

{% for r in result %}
    {% for k,v in r.items %}
        {{ k }} {{ v }}
    {% endfor %}j
{% endfor %}

you will then get:

opel 3
fiat 2
ford 2

Upvotes: 1

iMom0
iMom0

Reputation: 12911

I don't think dictsort is fit here.

>>> l = [('panda', 'fiat'), ('500', 'fiat'), ('focus', 'ford')]
>>> values = [i[1] for i in l]
>>> from collections import Counter
>>> c = Counter(values)
>>> c.most_common()
[('fiat', 2), ('ford', 1)]

You can define your own template tags,

from django import template
from collections import Counter

register = template.Library()

@register.filter
def my_sort(l):
    values = [i[1] for i in l]
    c = Counter(values)
    return c.most_common()

Upvotes: 2

Related Questions