SQL累计总和

Question

I am working with SQL Server and Management Studio 2014 and am looking to calculate cumulative dosing of a drug over time for each case in a database. A hypothetical sample of cases and times follows.

Case_Name    | InductionTime           | DoseTime                | EndTime                 |NEdose
2Z378J01     | 2007-03-07 08:58:00.000 | 2007-03-07 09:35:00.000 | 2007-07-03 13:06:15.000 | 0.01
2Z378J01     | 2007-03-07 08:58:00.000 | 2007-03-07 12:04:00.000 | 2007-07-03 13:06:15.000 | 0.02
2Z378J01     | 2007-03-07 08:58:00.000 | 2007-03-07 12:53:00.000 | 2007-07-03 13:06:15.000 | 0
2Z668I02     | 2007-06-06 08:46:00.000 | 2007-06-06 13:22:00.000 | 2007-06-06 13:51:45.000 | 0.04
2Z9C8E01     | 2007-09-12 08:46:00.000 | 2007-09-12 11:22:00.000 | 2007-12-09 11:46:00.000 | 0.02
2Z9C8E01     | 2007-09-12 08:46:00.000 | 2007-09-12 11:30:00.000 | 2007-12-09 11:46:00.000 | 0
2Z9C8E01     | 2007-09-12 08:46:00.000 | 2007-09-12 11:42:00.000 | 2007-12-09 11:46:00.000 | 0.01
2Z9C8E01     | 2007-09-12 08:46:00.000 | 2007-09-12 11:44:00.000 | 2007-12-09 11:46:00.000 | 0.02

What I need is to find the sum of the following:

[NEdose1] x ([InductionTime] - [DoseTime1]) +      //only if [InductionTime] > [DoseTime1]
[NEdose1] x ([DoseTime2] - [DoseTime1]) +
[NEdose2] x ([DoseTime3] - [DoseTime2]) + ... +
[NEdose(n-1)] x ([DoseTime(n]) - [DoseTime(n-1)])
[NEdose(n)] x ([EndTime] - [DoseTime(n)])  = TotalDose

Where [NEdose1] represents the value of the NEdose column in row 1 when the table is ordered by DoseTime. 

The numbering scheme is the same for DoseTimes continuing to row n

Grouped by each Case_Name. Some of the cases have only a NEdose and DoseTime, some have up to 20. The NE dose is in dose/minute, so the units of the time subtraction would be in minutes.

If you have any suggestions on getting me started (I am just a "hobbyist" with SQL), they would be greatly appreciated.

Edit: I have tried to clarify my desired calculation

Answer 1

This is too long for a comment.

I don't follow your logic. You have expressions like NEdose1 in the description of the calculation, but there is no column with that name.

Given the data structure, I would expect dosetime , endtime , and nedose to be the key columns. Assuming that the dose is per day, something like:

select case_name,
       sum(nedose * (datediff(day, dosetime, endtime) + 1)) as totaldose
from t
group by case_name;

Given that you want a summary per case, this doesn't seem cumulative.