sql总和在多列中重复

Question

I'm trying to sum values from duplicates rows (with the same ID, Month and Person) in multiple columns to the first or the last duplicate row. Then delete the duplicate rows exept the one with the total value. The biggest problem is that sometimes I need to sum values in two different columns.

PrimaryTable:

  ID Month Person  Value1 Value2
**123  1   Smith**   10     20  
**123  1   Smith**   5      NULL
**123  1   Smith**   NULL   5
  123  2   Smith     10     5
**189  3   Murphy**  NULL   15 
**189  3   Murphy**  NULL   10 
  190  2   Brown     25     25
**345  2   Lee**     25     20 
**345  2   Lee**     25     20 

Result1 (expected result after sum duplicates values to the first one):

ID Month Person Value1 Value2
123  1    Smith **15** **25** 
123  1    Smith   5      NULL
123  1    Smith   NULL   5
123  2    Smith   10     5 
189  3    Murphy  NULL **25** 
189  3    Murphy  NULL   10
190  2    Brown   25     25
345  2    Lee   **50** **40**
345  2    Lee     25     20 

FinalTable (expected result after deleting duplicates, except the first one):

ID Month Person Value1 Value2
123  1    Smith **15** **25** 
123  2    Smith   10     5 
189  3    Murphy  NULL **25** 
190  2    Brown   25     25
345  2    Lee   **50** **40** 

I'm trying with this code:

SELECT ID, Month, Person, SUM(Value1), SumValue2
FROM
(
    SELECT ID, Month, Person, Value1, SUM(Value2) AS SumValue2
    FROM db.Hours
    GROUP BY ID, Month, Person, Value1
 )
 GROUP BY ID, Month, Person, SumValue2

But sometimes it makes double sum of total of Value2.

Answer 1

SELECT ID, Month, Person, SUM(Value1) as SumValue1, SUM(Value2) AS SumValue2
FROM db.Hours
GROUP BY ID, Month, Person

I am not sure why you are looking at this as two steps etc. There is no removal of duplicates etc. this is a scenario for Group By Aggregation. Where you group like rows and summarize the value columns. The only reason you would need to make this a multi step operation would be if one of your value columns will be considered within your grouping eg ID, Month, Person, and Value1. In your case you simply need to group by ID, Month, Person and do the aggregation for Value1 and Value2.