在SQL (PostgreSQL)中，根据最接近日期值的其他字段的唯一组合对字段求和。

SELECT *
FROM your_table
ORDER BY date_column;

SELECT *,
       date_column - LAG(date_column) OVER (ORDER BY date_column) AS date_diff
FROM your_table
ORDER BY date_column;

SELECT *,
       ROW_NUMBER() OVER (ORDER BY date_column) AS row_num,
       date_column - LAG(date_column) OVER (ORDER BY date_column) AS date_diff
FROM your_table
ORDER BY date_column;

SELECT *,
       ROW_NUMBER() OVER (PARTITION BY row_num ORDER BY date_diff) AS diff_row_num,
       date_column - LAG(date_column) OVER (ORDER BY date_column) AS date_diff
FROM (
    SELECT *,
           ROW_NUMBER() OVER (ORDER BY date_column) AS row_num,
           date_column - LAG(date_column) OVER (ORDER BY date_column) AS date_diff
    FROM your_table
    ORDER BY date_column
) subquery;

SELECT SUM(your_field) AS sum_field
FROM (
    SELECT *,
           ROW_NUMBER() OVER (PARTITION BY row_num ORDER BY date_diff) AS diff_row_num,
           date_column - LAG(date_column) OVER (ORDER BY date_column) AS date_diff
    FROM (
        SELECT *,
               ROW_NUMBER() OVER (ORDER BY date_column) AS row_num,
               date_column - LAG(date_column) OVER (ORDER BY date_column) AS date_diff
        FROM your_table
        ORDER BY date_column
    ) subquery
) subquery2
WHERE diff_row_num = 1;