Rank top 3 with 2 dimensions

[#1]

Hello,

I want the top 3 theft of cars by 2 dimensions city and by brand of cars. So for example this is my table:

city (dimension)	brand car (dimension)	car theft (expression)
Groningen	toyota	5%
	volvo	14%
	bmw	18%
Amsterdam	toyota	2%
	renault	7%
	mercedes	9%
Rotterdam	volkswagen	31%
	citroen	9%
	peugeot	22%

Then i want to show only rotterdam Volkswagen (31%), Rotterdam Peugeot(22%) and Groningen bmw(18%). I got already a document with the title ' show top  performers'. This document is telling about how you can rank with the formula rank and agr. I tried a lot of things with this. But i do not get it right. Does somebody know how you can do something like this?

regards,

Camiel

[swuehl]

That's because in your set expression for the increase calculation, you filter for Brands with more than 5 thefts, not for the combination Brand&City. To enable this, create a field in the data model

LOAD

    City,

    Brand,

    Autonumber(City & Brand) as CityBrandID,

    ..

And use this in your set analysis.

See attached

[#3]

Thank you very much! This is great!

[#4]

Hello Swuehl,

i got it almost in my dataset. i got the top 3. But i still see the others. See the attached file. The formula looks like the same as you. What am i doing wrong?

[swuehl]

You just need to edit your chart, in chart properties, go to something called Add-On, data processing (? not sure, translated it from my language), then disable 'show NULL values'.

[#6]

I got it! thank you very much!