Question

I have a database that looks like the following;

circuit_uid   |  customer_name   | location      | reading_date | reading_time | amps | volts  |  kw  | kwh | kva  |  pf  |  key
--------------------------------------------------------------------------------------------------------------------------------------
cu1.cb1.r1    | Customer 1       | 12.01.a1      | 2012-01-02   | 00:01:01     | 4.51 | 229.32 | 1.03 |  87 | 1.03 | 0.85 |    15
cu1.cb1.r1    | Customer 1       | 12.01.a1      | 2012-01-02   | 01:01:01     | 4.18 | 230.3 | 0.96 |  90 | 0.96 | 0.84 |    16
cu1.cb1.s2    | Customer 2       | 10.01.a1      | 2012-01-02   | 00:01:01     | 7.34 | 228.14 | 1.67 | 179 | 1.67 | 0.88 | 24009
cu1.cb1.s2    | Customer 2       | 10.01.a1      | 2012-01-02   | 01:01:01     | 9.07 |  228.4 | 2.07 | 182 | 2.07 | 0.85 | 24010
cu1.cb1.r1    | Customer 3       | 01.01.a1      | 2012-01-02   | 00:01:01     | 7.32 | 229.01 | 1.68 | 223 | 1.68 | 0.89 | 48003 
cu1.cb1.r1    | Customer 3       | 01.01.a1      | 2012-01-02   | 01:01:01     | 6.61 | 228.29 | 1.51 | 226 | 1.51 | 0.88 | 48004

What I am trying to do is produce a result that has the KWH reading for each customer from the earliest (min(reading_time)) on that date, the date will be selected by the user in a web form.

The result would be/should be similar to;

Customer 1   87
Customer 2   179
Customer 3   223

There are more than the number of rows per day shown here and there are more customers and the number of customers would change regularly.

I do not have much experience with SQL, I have looked at subqueries etc. but I do not have the chops to figure out how arrange it by the earliest reading per customer and then just output the kwh column.

This is running in PostgreSQL 8.4 on Redhat/CentOS.

Answer 1

select customer_name,
       kwh,
       reading_date, 
       reading_time
from (
   select customer_name,
          kwh,
          reading_time,
          reading_date,
          row_number() over (partition by customer_name order by reading_time) as rn
   from readings
   where reading_date = date '2012-11-17'
) t
where rn = 1

As an alternative:

select r1.customer_name,
       r1.kwh, 
       r1.reading_date,
       r1.reading_time
from readings r1
where reading_date = date '2012-11-17'
and reading_time = (select min(r2.reading_time)
                    from readings
                    where r2.customer_name = r1.customer_name
                    and r2.read_date = r1.reading_date);

But I’d expect the first one to be faster.

Btw: why do you store date and time in two separate columns? Are you aware that this could be handled better with a timestamp column?