Whilst layering transparent charts is possible it tends to get very confusing very quickly - and sometimes the lining up fails.
You could use a combo chart and put a symbol above the bars to show the total (you would need to use the TOTAL <Dim> syntax to ignore the second dimension). Another way to achieve similar may be to have a stacked bar, where you have the DimValue as one expression then the Total (using TOTAL again) less the dimension value as expression two then stack them. I'm not sure whether stacked bars and multi dimensions play nicely together though.
It's always worth trying to push back to the end user to get them to have something a bit more standard, if you can.
I have edited Manish's Chart a little it and it looks exactly as what u want..
edited.qvw 143.2 K
Load * Inline
Month, Product, Sales
Jan, A, 200
Jan, B, 300
Jan, C, 120
Feb, A, 420
Feb, B, 360
Feb, C, 220
Mar, A, 180
Mar, B, 300
Mar, C, 420
Apr, A, 600
Apr, B, 400
Apr, C, 220
Create a bar cahrt with Month and Product as Dimension
In expression create two expressions
2. (TOTAL <Month> Sales)-sum(Sales)
for expression 1. background color
If(Product = 'A', Green(),
If(Product = 'B', Blue(),
If(Product = 'C', Red())))
In the color tab keep first three colors as light grey..
Great! Achieved after making bar distance Zero in Presentation tab.
One more thing! In expression tab, if i check 'values on data points' in first sum(sales) expression, it's giving correct. But if I check the same for 2nd TOTAL expression, it's giving three totals, though i need one. for example, in Jan month, I need 620 on top of gray bar.
Thanks a lot! dear.
I used Combo chart. First two expressions with bar. third expression with symbol (plus) with formula SUM(TOTAL<Month>Sales).
But if you have one product selected in the list, the lining up of two 'Values on Data points' was not good as pointed by stevedark.
However, to avoid lining up issue I made third expression conditional as =GetSelectedCount(Product)=0 or GetSelectedCount(Product)>1.
In the end, it all worked great.
Thanks a lot!