# Multi variable clustering cartographical data

Firstly , my knowledge in statistics is very limited , so excuse me if I'm ask a none well placed question.

I have a country ,( example : USA ) , and i have 3 set of data 1) position (lat , lng ) of every food storage (S) . 2) every storage have a variable C with , the C is the capacity . 3) a group of position (lat , lng) to specifing boards of every state

I need to cluster this data based on distance , capacity , and the border of any cluster need to be as similar to state boards

Ex : we need to create 3 k
The output will be : 33% of the capacities of all storages house in the first cluster , the second 33% , etc ... But that not mean the cluster will take 33% of the map , because that will gives an inequality between clusters . Note , the border alignment for states is used to to the same effect of a map matching .

What clustering algorithm i need to use ? And any idea of how i can create a visual image for this clustering ? Thanks

• Question is unclear, try to explain better. You need to create 3 k... what? What are the 33 %? – user2974951 Jan 7 at 10:34
• @user2974951 , yes , 3 k , k is number of cluster 3 k means 3 clusters , ( but its can be 4 , 10 , 1000 , ) it's just an example and about the 33% , if we have a 100% and 3 cluster , every cluster will have 33% ( is just an example again ) , if we have 4 cluster , every cluster will take 25% ... – MoxGeek Jan 7 at 14:22
• Have a look at stackoverflow.com/questions/8796682/…, if this is what you need. – user2974951 Jan 8 at 9:19