Hello,
Why is a large range of inaccuracies in trends_uint?
Sorry for my bad English.
Experience 27 Jun 11:00
1) Time from 22 Jun 2013 00:00 to 23 Jun 2013 00:00
Inaccuracy: -<1% trends 24 hours
Inaccuracy: -15,9% trends 24 hours !?
2) Time from 21 Jun 2013 23:00 to 22 Jun 2013 00:00
Inaccuracy: +2,1% trends 1 hour
Inaccuracy: -9,5% trends 1 hour
Can anyone verify this?
Why is a large range of inaccuracies in trends_uint?
Sorry for my bad English.
Experience 27 Jun 11:00
1) Time from 22 Jun 2013 00:00 to 23 Jun 2013 00:00
Code:
select AVG(value) AS value_avg from history_uint where itemid='23328' and clock>='1371844800' and clock<='1371931200';
value_avg
-----------------------
1438.8833333333333333
(1 row)
select AVG(value_avg) AS avg_value_avg from trends_uint where itemid='23328' and clock>='1371844800' and clock<='1371931200';
avg_value_avg
-----------------------
1426.8400000000000000
(1 row)
Code:
select AVG(value) AS value_avg, COUNT(value) AS count from history_uint where itemid='23329' and clock>='1371844800' and clock<='1371931200';
value_avg | count
---------------------+-------
63.1777777777777778 | 1440
(1 row)
select AVG(value_avg) AS avg_value_avg, COUNT(value_avg) AS count from trends_uint where itemid='23329' and clock>='1371844800' and clock<='1371931200';
avg_value_avg | count
---------------------+-------
53.4800000000000000 | 25
(1 row)
2) Time from 21 Jun 2013 23:00 to 22 Jun 2013 00:00
Code:
select AVG(value) AS value_avg, COUNT(value) AS count from history_uint where itemid='23328' and clock>='1371841200' and clock<='1371844800';
value_avg | count
-----------------------+-------
1421.0666666666666667 | 60
(1 row)
select AVG(value_avg) AS avg_value_avg from trends_uint where itemid='23328' and clock>='1371841200' and clock<='1371844800';
avg_value_avg
-----------------------
1451.0000000000000000
(1 row)
Code:
select AVG(value) AS value_avg, COUNT(value) AS count from history_uint where itemid='23329' and clock>='1371841200' and clock<='1371844800';
value_avg | count
---------------------+-------
63.2000000000000000 | 60
(1 row)
select AVG(value_avg) AS avg_value_avg from trends_uint where itemid='23329' and clock>='1371841200' and clock<='1371844800';
avg_value_avg
---------------------
57.0000000000000000
(1 row)
select num, value_avg, clock from trends_uint where itemid='23329' and clock>='1371841200' and clock<='1371844800';
num | value_avg | clock
-----+-----------+------------
60 | 56 | 1371841200
60 | 58 | 1371844800
(2 rows)
Can anyone verify this?
Comment