Django：关于带查询的优化分组

Question

我有一个这样的模型：

class ConsumptionTracking(Model):
        cat = models.ForeignKey('Cat', null=True, db_index=True, blank=True)
        time = models.TimeField(null=False, blank=False)
        date = models.DateField(null=False, db_index=True, blank=False)
        water_amount = models.IntegerField(null=True, blank=True)
        food_amount = models.IntegerField(null=True, blank=True)

我想检索某个日期的所有对象，并按日期对它们进行分组，例如：

{"01-02-2016": [ConsumptionTracking<1>, ConsumptionTracking<2>], 
"02-02-2016": [ConsumptionTracking<3>, ConsumptionTracking<x>, ConsumptionTracking<x+1>]}

有数千个条目，我发现使用set和for循环的唯一方法大约需要2秒，这太多了：

model = ConsumptionTracking
user = request.user
data = set(i for i in model.objects.filter(cat=cat, cat__owner=user, date__gte=date_from))
date_set = set([str(o.date) for o in data])
aggregated = dict([(d, []) for d in date_set])
for d in date_set:
    for o in data:
        if str(o.date) == d:
            aggregated[d].append(ConsumptionTrackingSerializer(o).data)

我可以在QuerySet级别执行此操作吗？我能跑得更快吗

谢谢