Django的聚集跨反向关系 [英] Django Aggregation Across Reverse Relationship

问题描述

由于这两款车型：

 类档案（models.Model）：
    用户= models.ForeignKey（用户，独特= TRUE，verbose_name = _（'用户'））
    关于= models.TextField（_（关于），空白= TRUE）
    邮编= models.CharField（MAX_LENGTH = 10，verbose_name ='拉链code'，空白= TRUE）
    网站= models.URLField（_（'网站'），空白= TRUE，verify_exists = FALSE）类ProfileView（models.Model）：
    简介= models.ForeignKey（简介）
    观众= models.ForeignKey（用户，空白= TRUE，空=真）
    创建= models.DateTimeField（auto_now_add = TRUE）
 

我想通过总的观点排序的所有配置文件。我可以通过总的观点与分类轮廓ID列表：

<$p$p><$c$c>ProfileView.objects.values('profile').annotate(Count('profile')).order_by('-profile__count')

但是，这只是一个配置文件ID的字典，这意味着以后我就遍历它放在一起轮廓对象的列表。这是许多附加的查询和仍然不会导致一个查询集。在这一点上，我还不如下降到原始的SQL。我这样做之前，有没有办法从剖面模型做到这一点？ ProfileViews通过一个ForeignKey的领域有关，但好像剖面模型的人都知道它是没有，所以我不知道如何配合两个在一起。

顺便说一句，我知道我可以只保存看法的剖面模型的属性，并可能变成是什么，我在这里做的，但我仍然有兴趣学习如何更好地使用聚合函数。

解决方案

  

ProfileViews通过一个ForeignKey的领域有关，但好像剖面模型的知道它不是

配置文件模型实际上不知道。
你应该能够做到这样的：

  Profile.objects.all（）。注释（count_views = COUNT（'profileview__pk'））。ORDER_BY（'count_views'）
 

修改：在这里你可以找到有关跨越关系<一个查找Django文档href=\"http://docs.djangoproject.com/en/dev/topics/db/queries/#lookups-that-span-relationships\">http://docs.djangoproject.com/en/dev/topics/db/queries/#lookups-that-span-relationships

Given these two models:

class Profile(models.Model):
    user = models.ForeignKey(User, unique=True, verbose_name=_('user'))
    about = models.TextField(_('about'), blank=True)
    zip = models.CharField(max_length=10, verbose_name='zip code', blank=True)
    website = models.URLField(_('website'), blank=True, verify_exists=False)

class ProfileView(models.Model):
    profile = models.ForeignKey(Profile)
    viewer = models.ForeignKey(User, blank=True, null=True)
    created = models.DateTimeField(auto_now_add=True)

I want to get all profiles sorted by total views. I can get a list of profile ids sorted by total views with:

ProfileView.objects.values('profile').annotate(Count('profile')).order_by('-profile__count')

But that's just a dictionary of profile ids, which means I then have to loop over it and put together a list of profile objects. Which is a number of additional queries and still doesn't result in a QuerySet. At that point, I might as well drop to raw SQL. Before I do, is there a way to do this from the Profile model? ProfileViews are related via a ForeignKey field, but it's not as though the Profile model knows that, so I'm not sure how to tie the two together.

As an aside, I realize I could just store views as a property on the Profile model and that may turn out to be what I do here, but I'm still interested in learning how to better use the Aggregation functions.

解决方案

ProfileViews are related via a ForeignKey field, but it's not as though the Profile model knows that

The Profile model actually does know that. You should be able to do something like:

  Profile.objects.all().annotate(count_views=Count('profileview__pk')).order_by('count_views')

Edit: Here you can find Django docs about lookups that span relationships http://docs.djangoproject.com/en/dev/topics/db/queries/#lookups-that-span-relationships

这篇关于Django的聚集跨反向关系的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！