Jump to content


Classifying spatial data

- - - - -

  • Please log in to reply
1 reply to this topic



    Key Contributor

  • Validated Member
  • PipPipPip
  • 62 posts
  • Norway

Hi all,

I am working on a project where i need to classify some variables.
My problem is that i am unsure of which classification method that is appropriate for my variables.
The variables are a population grid per cell for Africa, and the other is infant mortality rate per cell for Africa.
I was thinking about using natural breaks, but i think this method at least for infant mortality rate would remove the variation within some of the classes.

Do any people here have some suggestions on what is an apropriate way of classifying these variables, and why? I would need to explain my choice.
Lets say 5-7 classes for each variable.

Visit my Blog: http://www.gisintersect.com



    Hall of Fame

  • Admin
  • PipPipPipPipPipPipPip
  • 2,420 posts
  • Gender:Male
  • Location:Stockholm, Sweden
  • Interests:music, hiking, friends, nature, photography, traveling. and maps!
  • Sweden

This is for presentation, right? Then I would think about what messages you want to present, and also see what the distribution looks like. For these variables - also keep in mind that they are best estimates/approximations/modeled values, so I would keep these in mind when creating the classification. For presentation, I usually try to have rounded class breaks.
Hugo Ahlenius
Nordpil - custom maps and GIS

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users