Django way to do double annotation

Using Django Using the ORM
1

Hey, I have a model LeadsClicks
I create from this model every time someone goes to one of my website links

class LeadsClicks(models.Model):
    created_at = models.DateTimeField(_("created at"), auto_now_add=True)
    business = models.ForeignKey(
        Business,
        on_delete=models.CASCADE,
        related_name="leads_clicks",
        verbose_name=_("business"),
    )
    qr = models.ForeignKey(
        "BusinessQR",
        on_delete=models.CASCADE,
        related_name="leads_clicks",
        null=True,
        blank=True,
        verbose_name=_("QR"),
    )
    referrer = models.CharField(_("referrer"), max_length=1000, blank=True, null=True)

I want to create an api and return results summed daily

"results": [
        {
            "day": "2025-06-22",
            "clicks": 63,
            "info": {
                "qr__name1": 60,
                "qr__name2": 3
            }
        },
        {
            "day": "2025-06-29",
            "clicks": 51,
            "info": {
                "qr__name1": 46,
                "qr__name2": 1,
                "qr__name3": 2,
                "qr__name4": 2
            }
        },

I manage to do it, but not 100% in the Django DRF way…
How could I improve it? Ordering is done manually, and everything feels wrong…

class DailyLeadsClicksAPIView(viewsets.ReadOnlyModelViewSet):
    #serializer_class = LeadsClicksSerializer2
    pagination_class = CustomPagination
    filter_backends = [filters.OrderingFilter, DjangoFilterBackend]
    filterset_class = LeadsClicksCustomFilter
    ordering_fields = ['day', 'clicks']
    ordering = ['-day']

    def get_queryset(self):
        biz = self.request.user.profile.biz

        qs = LeadsClicks.objects.filter(business=biz, qr__isnull=False).annotate(
            day=TruncDate('created_at')
        ).annotate(
            clicks=Count('id', distinct=True)
        ).values('day', 'qr__name', 'clicks')
        return qs

    def list(self, request, *args, **kwargs):
        # from queryset
        # (day, qr__name, clicks)
        # to resp:
        # (day, clicks, info: {key: qr__name, value: clicks})
        queryset = self.filter_queryset(self.get_queryset())

        page = self.paginate_queryset(queryset)
        
        
        from collections import defaultdict
        result = defaultdict(lambda: {"clicks": 0, "info": {}})
        for row in queryset:
            day_str = row['day'].isoformat()
            result[day_str]["clicks"] += row['clicks']
            if result[day_str]["info"].get(row['qr__name']):
                result[day_str]["info"][row['qr__name']] += row['clicks']
            else:
                result[day_str]["info"][row['qr__name']] = row['clicks']

        resp_data = [
            {"day": day, "clicks": data["clicks"], "info": data["info"]}
            for day, data in result.items()
        ]
        
        # Sort by clicks manually
        ordering = request.query_params.get('ordering')
        if ordering == 'clicks':
            resp_data.sort(key=lambda x: x['clicks'], reverse=False)
        elif ordering == '-clicks':
            resp_data.sort(key=lambda x: x['clicks'], reverse=True)
        elif ordering == '-day':
            resp_data.sort(key=lambda x: x['day'], reverse=True)
        elif ordering == 'day':
            resp_data.sort(key=lambda x: x['day'])

        page_number = int(self.paginator.get_page_number(request, self.paginator)) - 1
        page_size = int(self.paginator.get_page_size(request))
        resp_data2 = resp_data[page_size*page_number: page_size*(page_number+1)]  
        resp = self.get_paginated_response(resp_data2)
        resp.data['count'] = len(resp_data)
        return resp

BTW: The info can have a different shape as appropriate. And I use PostgreSQL

I did not find a way to do it with a single query and without Python