Comments by Petri Luoto

Petri Luoto 13-Jan-12 14:55pm View

Processing with real data 780.000 lines resulting in 268.000 lines took 11 seconds! Great! Also the solution is simple and the reulting table is clear with original data types:
2012-01-06 09:01 3581
2012-01-06 09:00 3222
2012-01-06 08:59 3268
2012-01-06 08:58 3314
2012-01-06 08:57 3254

I will use this solution. Thanks!

Petri Luoto 13-Jan-12 14:49pm View

Thanks! Processing of 780.000 lines resulting in 268.000 lines took 14 seconds. Quite nice. But there was a problem: the time column went funny after 59 minutes:

2012-01-06 00:00:00.000 9:1:00 3581
2012-01-06 00:00:00.000 9:0:00 3222
2012-01-06 00:00:00.000 8:9:00 1116
2012-01-06 00:00:00.000 8:8:00 944
2012-01-06 00:00:00.000 8:7:00 960
2012-01-06 00:00:00.000 8:6:00 960
2012-01-06 00:00:00.000 8:59:00 3268
2012-01-06 00:00:00.000 8:58:00 3314
2012-01-06 00:00:00.000 8:57:00 3254

Petri Luoto 13-Jan-12 14:30pm View

Thanks! I tried this, but for some reason got an error message: "Argument data type time is invalid for argument 1 of substring function."
I did not investigate it closer, but perhaps something with datatypes. Those were Date, Time(7), and int.

Petri Luoto 12-Jan-12 6:53am View

Taking a closer look to result, it allways returns an array of 60 values. But actually it should return an array with 1440 values. In the table there are some 7300 values per day, which is around 5 values per minute. Those are the values which should be calculated for one average value, total of 1440 from the whole Table.

Perhaps you figure out another proposal, and with Time value included? Thanks in advance.

Comments by Petri Luoto (Top 4 by date)