Replies: 1 comment 1 reply
-
CueObserve assumes the last data point is partial and ignores it. In case of hourly anomalies, it should ignore the last hour data, not the last day's data. This is a known issue fixed by #179. Which version are you using? |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I am running the cueobserve for a dataset with hourly granularity but it does not seems to work on the current day data. Like If I am running the anomaly_definition today and giving the data till current hour, it is only calculating the anomalies till yesterday data.
Here is data I am Providing:
Here Is my query:
SELECT
TIMESTAMP_TRUNC(created_on, HOUR) as created_on,
status, count(*) as count
from x.y
where status in('SENT_TO_PROVIDER', 'DELIVERED')
group by 1,2
order by 1 desc
Here Is the Anomaly Graph:
In this graph anomalies is calculated only till last day.. for the current day it is creating the dashed-line graph.
In my use case I wanted to check for the anomalies for current day data.
Is it something expected behaviour of Cueobserve or Am I making any mistake in query?
Beta Was this translation helpful? Give feedback.
All reactions