How to exclude repeating values while creating an Area Vizualization and using Unique Count?

[iraidalavrovaa (Iraida)]

Hi all!

I want to visualize a trend of appearing of new users over time. Username is a text field. I decided to use Area Visualization with Cumulative Sum and Unique count of username aggregations. I specified it in Metrics. Then there was a hint to use Date Histogram or Histogram in aggregation in Buckets. So now I have a timestamp on the X-axis.

I see that the values on the graphs are too big because it makes the sum of unique users per periods and ignore intersections. I mean I want to make sums of unique users from the last period and new (who appear for the first time) users.

Any ideas, please? Maybe JSON advanced script is needed?
Thank you!

[Sandeep_Raju (Sandy)]

As per your requirement, You want count of new users over a period of time...

For this , we can use Bar or Area chart, then use a date field like updated/created_time in x axis , then put metric as unique count of Usernames.

It will give total number of new usernames by time

I don't get why you should use sum for this....

[iraidalavrovaa (Iraida)]

Thank you for the idea. I'm trying to use cumulative sum because I'm interested in growth rate. For further analysis of periods when we had a lot of (or a few) new users

[Sandeep_Raju (Sandy)]

I think TSVB visualization is more helpful in this scenario...