So i am using django and get user's location at registration time.

Then i show these users on the front page of the app but sorted as per the distance, i.e, the closest ones to the logged in user are on the top and so on.

Now what i am doing is i am ordering them as per distance on the backend using some annotate (etc) functions provided by django ORM.

sortedQueryset = self.get_queryset().annotate(distance=Distance(
                'coords', user.coords, spheroid=True)).order_by('distance')

Where 'coords' is the column in db to store the point (location), user.coords is point (coordinates) of the logged in user.

Now to get only first 100 users (say) from the database i can do something like this;

sortedQueryset = self.get_queryset().annotate(distance=Distance(
                'coords', user.coords, spheroid=True)).order_by('distance')[:100]

But what it think, it still grabs all the rows, orders them as per distance and then gets 100 of them. Say we have a million users in db, then it always has to get all those and then sort them and then get only 100.

I think it is a lot of overwork (maybe i am wrong or maybe this is the only way as i have to sort as per distance and that also depends on the logged in user, who is closest and who is farthest).

Any suggestions are appreciated. Thanks!

CodePudding user response：

Actually what you have done is right only. This will not slice in Python but limit it in the database query itself. So it won't get all the results and slice it, instead, it runs LIMIT query against the database. See the documentation.

https://docs.djangoproject.com/en/dev/topics/db/queries/#limiting-querysets