Hi All,
I am wanting to create a top N chart. The top one is aggregated data by “Date”, and the bottoming is granular (by field “Name”) .
There is a parameter option to choose date granularity.
I have added a cross tab for count against timestamp and against date.
Whenever a user selects a date range and “names” from the drop down, the users should see the aggregated values in the top chart ( for all the “Names” selected) and for the date granularity selected.
The bottom chart should break the results down to a lower granularity by Names.
However,
I am getting erratic results.
For example when I choose Date granularity as “hours” and Top N parameter as 19, and choose multiple “Names” I am not getting the correct results. for a date range os 3/1 to 3/31 and for name “bb” I am only getting the Top 15 though I have selected top 19.
(If I select only one “Name” I am getting the correct result though. )
I have used table calcs and have also used contact filters.
I am attaching the dashbaord for your reference.
Any help would be appreciated.
Thank you!
Regards,
Rishi Saka
@Rishi Saka I didn't look at the Granular sheet previously, so I didn't notice that you have [Name] as a context filter as well. This means whenever you filter you are getting different sets (an option I mentioned in the last post).
I made a mistake with the screenshot in the last post. I didn't sort first on the 'bb' values, so while the values circled in red would be ranked higher than some of the values in the current set, none of them would be in a set of just 'bb'. In fact none of the hours in the Total/'aa' set are in the 'bb' set.
You have to determine how you want the two dashboards to relate
- The top dashboard shows the Top N Total count in chosen time period. The bottom dashboard shows the breakdown by name of those hours. In this case. remove [Name] from filter.
- The top dashboard shows the Top N Total count in chosen time period. The bottom dashboard show the Top N count by name. The top time periods may differ for each group (total, 'aa', 'bb') In this case, create three different sets instead of using one set for all graphs.
The screenshot illustrates the difference (I had to reduce 19 to 5 so the annotations/labels can be seen). The top graph is the the Total Count set. The second graph is the 'bb' members of that Total set. There are only 4 values because one hour of the Total set does not have a 'bb'. For the second graph, [Name] is not on the filter shelf. The third graph is the top 5 'bb' values. The day/hours in this set are completely different than those in the Top 5 Total set. For this graph, [Name] is a context filter.
Because sets are involved, the only two meaningful options are no filter or context filter. Using a standard dimension filter would behave the same as no filter.
