I normally don't use the key value object due to the problems you mention. I use the text object instead where I have more control 🙂 You can add text and measure/parameter values and then format them to your liking.   BR Fredrik ... View more

Hi! I think it only works if the two measures have the same format. When I use your method i usualy try to make it very clear what is shown in the graph. In some reports I added a textobject "on the graph" showing the parameter value 🙂   Another way of solving this could be to use a stacked(tabbed)-container with one tab for value and another for %. I think data is loaded for each tab directly so it is very fast when changing tab. I use it a lot.     BR Fredrik   ... View more

One way I think might work is to create a new aggregated table containing year and count of number of observations. You might also add some dimensions if you filter on these in the report.   ... View more

Could you do that using tabs, either using the tab container och tabs in the report? ... View more

Hi! You can create something similiar by using a bar chart.   Create a calculated measure with value 1 and add to the bar chart as measure. Add your time dimension to the bar chart as categoy. group the bar chart (rows) on your machine (or what you use), as in yourexample grd_a and grd_b. Create calculated measures for each of your statAcronym Create a display rule for each of the measures. Could look something like this: BR Fredrik ... View more

You can use a data driven content and create a viya job that creates the table dynamically and display it in VA. Might help you...:) BR Fredrik ... View more

I think you are on the right track by creating different dimension variables depending on what the "General RTA category" contains. The visualization part can be done by using different objects for each of the variables. Maybe different bar charts, key indicators can be used? Another way could be to create differnt meassures depending on the contined value and then use these in an bar chart as different measures. In that case tou can create a "total" dimension variable with only one value; "Total". Just my 5 cents, have not tested my thoughts 🙂 ... View more

Could it work by create numeric variables of the dates and then use the aggregate table function to get min/max per id, and then calculate the days? BR Fredrik ... View more

Hi! I don't think it is possible without pre-arranging the data. I usually join the interval-data with a calendar-dataset before loading it to VA, but the data grow very quick 🙂 //Fredrik ... View more

Sorry for late answer, you might already found a fix, but anyways:   Create a textvariable for weekday: IF ( DayOfWeek('date'n) = 1 ) RETURN 'Sunday' ELSE ( IF ( DayOfWeek('date'n) = 2 ) RETURN 'Monday' ELSE ( IF ( DayOfWeek('date'n) = 3 ) RETURN 'Tuesday' ELSE ( IF ( DayOfWeek('date'n) = 4 ) RETURN 'Wednesday' ELSE ( IF ( DayOfWeek('date'n) = 5 ) RETURN 'Thursday' ELSE ( IF ( DayOfWeek('date'n) = 6 ) RETURN 'Friday' ELSE 'Saturday' ) ) ) ) )   and one for date: Format('date'n[Raw], 'YYMMDD8.')   And use them instead of the date variables, should work 🙂   //Fredrik ... View more

I think the date format is the problem. Is it possible to create two text objects containing day/week and then use them and the format string? //Fredrik ... View more

What do you want to achieve? How would you like to have it visualized? //Fredrik ... View more