Design Question: Low load data aggregation for overview page - django

Question

What is the best way to achieve low load on the database or application server for this use case:
Let's say I want to build a web application that has for each user an overview page. The overview page shows in an aggregated form for each user the user's data. For example, if it were a library application it would show how many times the user visited the library in total, how many books he read in total, how many books were delivered delayed in total, how many minutes he spend in the building. Each time the user visits the overview page the up-to-date values should be displayed. While the user interacts with the site the numbers change.
What I could do is for every overview page refresh do several counts in the database. But that would be expensive.
views.py
def overview(request, userID):
booksCount = Book.objects.count()
booksReadCount = Book.objects.filter(UserID=userID, Status='read').count()
# ... many more, same way
libraryVisitedCount = LibraryVisits.objects.filter(UserID=userID).count()
# many counts like these on different tables for the user
data = {
"booksCount" : booksCount,
"booksReadCount" : booksReadCount,
# ... many more
"libraryVisitedCount" : libraryVisitedCount
}
render(..., context=data)
I have thought I could store a JSON object with the data to be presented on the overview page in a database table and I update the JSON each time an event happend on the site which affects the count of objects.
Or I could use a materiliazed view but to refresh it I would have to recalculate all the data of all users each time, right?
Other ideas? I'm using django webframework and postgres database.
TL;DR: I wondered isn't there a better way to receive counts than do several counts in the database each time?
Thanks.

Answer 1

Lets say, in Book, LibraryVisit etc models, there is ForeignKey to User model with related_name like this:
class Book(models.Model):
UserID = models.ForeignKey(User, related_name='books', on_delete=DO_NOTHING)
class LibraryVisit(models.Model):
UserID = models.ForeignKey(User, related_name='library_visit', on_delete=DO_NOTHING)
Then you can use annotation and conditional expression like this:
from django.db.models import Case, IntegerField, Sum, When
def overview(request, userID):
users = User.objects.filter(pk=userId)
users = users.annotate(
booksReadCount=Sum(
Case(
When(book__Status='read', then=1),
output_field=IntegerField()
)
)
).annotate(library_visited_count=Count('library_visit'))
# FYI: please use snake_case when defining object attribute(like model fields) as per PEP-8 style guide
data = {
"user_object" : users.first(), # taking first item of the User queryset. Also DB is hit once in this step
"booksCount" : Book.objects.count()
}
# access counts in view like this:
# user.book_read_count
# user.library_visited_count
return render(..., context=data)
# bold marked words are related_name
And render counts in template like this:
{{ user_object.book_read_count }}
{{ user_object.library_visited_count }}