關於GROUP BY在Django

Question

我有一些問題：

# models.py 
class someinfo(models.Model): 
    data1 = models.BigIntegerField(verbose_name='data1') 
    data2 = models.BigIntegerField(verbose_name='data2') 
    data_pubdate = models.DateTimeField(default=timezone.now, verbose_name='data_pubdate') 
    people = models.ForeignKey(Drive, on_delete=models.CASCADE, verbose_name='people') 

數據表有：

# row data 
data1 data2  data_pubdate    people 
7777129 23232  2017-05-18 05:04:08.377  9 
158160 4573279  2017-05-18 05:04:08.737  8 
158200 5709307  2017-05-18 05:04:08.787  6 
10073 11612  2017-05-18 05:04:19.727  6 
2458  5245466  2017-05-18 05:04:19.727  4 
2458  5245466  2017-05-18 05:04:19.724  4 
14879 9727   2017-05-18 05:04:19.897  7 
...etc 

我要從新data_pubdate在每個人的選擇data1和data2的，像這樣的表：

# select data 
data1 data2  data_pubdate    people 
7777129 23232  2017-05-18 05:04:08.377  9 
158160 4573279  2017-05-18 05:04:08.737  8 
10073 11612  2017-05-18 05:04:19.727  6 
2458  5245466  2017-05-18 05:04:19.727  4 
14879 9727   2017-05-18 05:04:19.897  7 

那麼...我該怎麼辦？

P.S. Django的版本是 「1.10」，數據庫是 「MSSQL Server 2012的」

Answer 1

也許最簡單的方法是：

SomeInfo.objects.annotate(
    last_pubdate=Max('people__someinfo__data_pubdate') 
).filter(last_pubdate=F('data_pubdate'))