按特定时间间隔的日期时间分组

我有以下类别的对象列表：

public class CounterData 
{
    public DateTime counterTime { get; set; }
    public int counterName { get; set; }
    public int count { get; set; }
}

例如，我有以下格式的数据列表{counterTime, counterName, count}:

{"Aug  8 2016  9:00AM","counter1",11}
{"Aug  8 2016  9:05AM","counter2",12}
{"Aug  8 2016  9:11AM","counter3",47}
{"Aug  8 2016  9:12AM","counter3",20}
{"Aug  8 2016  9:13AM","counter1",12}
{"Aug  8 2016  9:30AM","counter3",61}
{"Aug  8 2016  9:35AM","counter2",35}
{"Aug  8 2016  9:39AM","counter1",16}
{"Aug  8 2016  9:40AM","counter1",92}
{"Aug  8 2016  9:53AM","counter2",19}

我想按以下方式对计数器进行分组counterName and counterTime每 15 分钟间隔聚合一次。对于上面的例子，结果列表应该是：

{"Aug  8 2016  9:00AM","counter1",23}
{"Aug  8 2016  9:00AM","counter2",12}
{"Aug  8 2016  9:00AM","counter3",67}
{"Aug  8 2016  9:30AM","counter3",61}
{"Aug  8 2016  9:30AM","counter2",35}
{"Aug  8 2016  9:30AM","counter1",108}
{"Aug  8 2016  9:45AM","counter2",19}

9:00AM-9:15AM，共有 2 条条目counter1。所以count值是条目的总和。其他计数器也类似。

我们可以使用LINQ 分组依据 https://msdn.microsoft.com/en-us/library/bb534304(v=vs.110).aspx解决这个问题？如果是这样那怎么办？

这是一个与时俱进的解决方案down精确到 15 分钟。它用AddSeconds基本上摆脱了秒部分DateTime, then AddMinutes with % https://msdn.microsoft.com/en-us/library/0w4e0fzs.aspx对分钟进行四舍五入：

var data = counters
    .GroupBy(cd => new
    {
        Date = cd.counterTime.AddSeconds(-cd.counterTime.Second)
                   .AddMinutes(-cd.counterTime.Minute % 15),
        CounterName = cd.counterName
    })
    .Select(g => new
    {
        Date = g.Key.Date,
        CounterName = g.Key.CounterName,
        SumCount = g.Sum(cd => cd.count)
    });

但是，这不适用于 LINQ to Entities（即实体框架），因此您需要稍微调整它才能使用DbFunctions方法：

var data = counters
    .GroupBy(cd => new
    {
        Date = DbFunctions.AddMinutes(
            DbFunctions.AddSeconds(cd.counterTime, -cd.counterTime.Second), 
            -cd.counterTime.Minute % 15),
        CounterName = cd.counterName
    })
    .Select(g => new
    {
        Date = g.Key.Date,
        CounterName = g.Key.CounterName,
        SumCount = g.Sum(cd => cd.count)
    });