Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Need help.
I am trying to generate a line chart in QlikSense Desktop to show counts by grouping a time stamp field with the help of expression.
The time stamp field used in the graph (END_TS_HR) is a computed field created using expression- Timestamp(END_TIME_STAMP,'YYYY-MM-DD hh:mm').
I am using below Aggr function to group STATISTIC_VALUE by END_TS_HR for the rows where STATISTIC_NAME = 'NumOrdersCreated'.
But the graph seems to shows END_TS_HR for all rows, irrespective of the value of STATISTIC_NAME.
Also the STATISTIC_VALUE is not getting grouped by END_TS_HR.
Dimension: END_TS_HR
Measure: Aggr(Sum({$<STATISTIC_NAME={'NumOrdersCreated'}>} STATISTIC_VALUE), END_TS_HR)
Expected result is -
END_TS_HR SUM
2018-08-16 00:00:00 36
2018-08-16 00:10:00 11
...
Update:
I was able to get rid of extra END_TS_HR that had no value, by un-checking Include Zero Values.
But I still have the challenge to group the STATISTIC_VALUE by END_TS_HR.
Dimensions and Measures are still same as before.
Here is current line chart.
Hi
Don't use aggr
Same with sum without aggr should work
So with aggr you will count also when stat_name is not what you want
If you have only sum your filter will be correct.
Regards
You are right Olivier, removing Aggr from measure expression also works-
Sum({$<STATISTIC_NAME={'NumOrdersCreated'}>} STATISTIC_VALUE).
However, it is still not grouping the STATISTIC_VALUE by END_TS_HR.
There is surely a problem with timestamp
You could try with date(date#(
It works better(look at posts with timestamp)
Date(Date#(END_TIME_STAMP,'YYYY-MM-DD hh:mm')) expression makes the computed field END_TS_HR blank.
Below expression for computing END_TS_HR works, but still do not see grouping of STATISTIC_VALUE by END_TS_HR. Above Graph stays same.
Date(END_TIME_STAMP,'YYYY-MM-DD hh:mm')
timestamp() is only formatting the timestamps, not grouping them per minute. I think what you need is to round the timestamps to the nearest ten minutes.
Change your calculated dimension to timestamp(round(END_TIME_STAMP,(10/(24*60))),'YYYY-MM-DD hh:mm')
If you want to specify if the function should round up or down change round() to either ceil() pr floor()