>> items.aggregate((,), sum=(
>> 'field1',
>> 'field2',
>> ...
>> 'field20',
>> 'field21',
>> ), average=(
>> 'field1',
>> 'field2',
>> ...
>> 'field20',
>> 'field21',
>> ))
>
> well, in this extreme example, I would suggest you use a list:
> fields = [ f.name for f in items.model._meta.fields ]
> items.aggregate( sum=fields, average=fields, min=fields, max=fields )
>
> not that bad, is it?
A nice way to do this, and a good compromise on a clean syntax that is
also easy to do what I need.
> true, but only if you would want to aggregate by those fields, we
> could work around that by simply moving the grouping fields to a
> separate dictionary as well:
>
> [
> {
> 'grouped_by' : { 'owner' : XX },
> 'min' : { 'pay' : 100 },
> 'max' : { 'pay' : 101},
> },
> .....
> ]
Another excellent idea. Pleasantly simple yet a good way to encapsulate
the info while still removing the ambiguity of field-names that conflict
with aggregate-function names. I'm still mildly concerned about being
able to access the contents from within a template, wherein one could
access it by member notation suggested earlier:
stats.grouped_by.owner
stats.sum.pay
stats.max.pay
-tkc
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to django-users@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/django-users?hl=en
-~----------~----~----~----~------~----~------~--~---
