View Full Version : Graphs and the 4-color theorem


Jerrek
11-20-03, 09:02 PM
I have a little problem here. How would I go about to prove that any planar graph which does not contain a diamond (let a diamond be a graph consisting of two triangles intersecting in one edge) is 4-colorable? Without using the 4-color theorem. Any advice as to the approach to the solution would be greatly appreciated.

RheniumIodide
12-06-03, 05:33 AM
hmm, seems like you can just say since there are no 'diamonds', there certainly aren't any K5's (complete graphs on 5 vertices), since the diamond you've described is a subgraph of the K5. Since you never have 5 vertices that are all adjacent to each other, you'd never need 5 colours

In fact, you'd never even need 4 colours. With no diamonds, the graph should be 3-colorable, because the diamond thing is a subgraph of a K4 as well

I assume that, by diamond, you mean something like this: (X's are vertices)

'''''X
''''/''\
''/'''''\
X-----X
''\''''''/
''''\''/
'''''X

(ignore all the ''' ...that's just for formatting)