- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
ā04-08-2019 08:15 AM
There are two PA reporting widgets which are showing wrong numbers and I have been looking all community posts and online resources for days but I am not able to resolve my issues. Please help to identify what I missed.
Current customer is using London family instance.
Business Requirement:
A date field: Date delivered
A currency field: All cost
1. get quarterly number of records based on Date delivered
For instance, there are 5 records and their Date delivered fall within FY Q1; there are 7 records and their Date delivered fall within FY Q2; there are 10 records and their Date delivered fall within future FY Quarter 4. The widget should display column for FY Q1 as 5, FY Q2 as 7, FY Q3 as 0, and FY Q4 as 10.
I created an Indicator Source set the condition to Date delivered on this Fiscal Quarter. Its Valid for frequency in the default value Daily.
An indicator without a frequency selection and it complies with the indicator source's frequency. The indicator has below configurations.
- Unit: #
- No additional conditions
The widget for the quarterly number of records has below configurations.
- Type: Time Series
- Time series: By fiscal quarter SUM +
The results show all numbers of records sums not getting results the customer wants. If I changed the time series, the results will show the all FY daily results or big number of records. How can I do to get the right quarterly column chart?
2. Show quarterly sum of All cost
I used the same indicator source and created a new indicator with following configurations.
- Unit: $
- Aggregate: Sum
- Field: All cost
- No additional conditions
The widget for the requirement has below configurations.
- Type: Time Series
- Time series: By fiscal quarter SUM +
The widget renders big expense. I also changed Time series but not getting correct results. How can I get the correct results?
Thank you in advance.
Solved! Go to Solution.
- Labels:
-
Performance Analytics
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
ā04-16-2019 08:25 AM
You indicator source should be something like Date delivered on this Fiscal Quarter AND Created on Yesterday. By excluding the condition after the AND, you're basically double counting. Instead of yesterday, you could use today, depending on when you run the data collection job. You may need to create a new indicator source instead of changing the existing one if there are other PA objects depending on the existing indicator source (since you are in London, you can use Dependency Assessment to analyze this).
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
ā04-08-2019 01:31 PM
What is the frequency of the indicator source? If your result is too large, it may be your indicator is daily when it should be quarterly. Or the indicator could be monthly and the condition should be monthly to match. If they are mismatched you would have issues.
Can you include screenshots of the indicator source and indicators?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
ā04-09-2019 02:49 PM
I left the default frequency alone. It shows Daily.
I have to build out new indicator source and indicators since time series calculation generates big numbers. If I did not choose any time series, it displays quarterly records but does not go far enough. I need to pick up future FY quarterly records as well. Let me work on my idea based on your input (frequency issue) and I will post my results.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
ā04-10-2019 06:59 AM
Hi yltsai,
I think the issue here is that source frequency, source conditions, and time series are not clearly understood.
Keep in mind that an indicator's value, or score, is a number that is collected by a regularly scheduled job.
In indicator source's conditions determine which records are included in the calculation when this number is collected. In your case you are instructing it to use all records in the current fiscal quarter. Every time a score is collected, it will use those conditions.
The indicator source's frequency determines how many scores the system will store. If it is daily, it will store one score/value per day. If it is monthly, it will store one value per month, and so on. You have opted for daily, so you get one score per day. But remember, that your condition is looking at the whole fiscal quarter. The result is that you are capturing a value for the current fiscal quarter for each day in the quarter. Whether that is what you want, I can't say. But, that's what you are getting. The value of this particular configuration is that you can trend how this score for the current quarter changes as the quarter progresses. If you are trying to achieve a target, this will show you whether you are on track to achieve it.
The last piece at play here is the time series. Time series allow you to aggregate an indicator's scores after they have been collected. You are using the "fiscal quarter SUM" series. This takes all the collected scores for each fiscal quarter, adds them up, and gives you the result for each quarter. Since you are basically collecting a running total for the quarter each day, applying this time series is adding these values up and giving you an inflated value.
Your next step is to determine what the actual requirement is (running total for current quarter with or without history) and adjust the frequency, conditions, and time series accordingly.
Best of luck!
Regards,
Jason Monzo
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
ā04-16-2019 08:25 AM
You indicator source should be something like Date delivered on this Fiscal Quarter AND Created on Yesterday. By excluding the condition after the AND, you're basically double counting. Instead of yesterday, you could use today, depending on when you run the data collection job. You may need to create a new indicator source instead of changing the existing one if there are other PA objects depending on the existing indicator source (since you are in London, you can use Dependency Assessment to analyze this).