Lattice Points inside Circles


By Maria Jose Leon-Sotelo Esteban on Friday, July 06, 2001 - 07:37 pm:

How can I obtain the formula for the number of points with integer coordinates inside a circle? Do you know some web page about this topic?
Thank you.
Maria Jose.


By Arun Iyer on Saturday, July 07, 2001 - 06:58 pm:

Maria,
say the center of the circle is (h,k) and let r be its radius.

Then the equation of the circle is
(x-h)2 + (y-k)2 = r2

If we could find all the integral solution to,
(x-h)2 + (y-k)2 < r2

then you would get your answer.
Any further suggestions guys?
love arun


By David Loeffler on Monday, July 09, 2001 - 01:01 am:

Must there necessarily be a neat formula? I think there was once a problem used in training for the William Lowell Putnam Contest (a very hard olympiad-type competition for American universtiy students) recently: show that for all n, there exists a circle in the plane containing precisely n points. That suggests that the general field is likely to be quite hard and there is unlikely to be a neat general formula.

P.S. See if you can solve the Putnam problem I mentioned! There's a very neat trick.


By Michael Doré on Monday, July 09, 2001 - 09:02 pm:

In the Putnam Contest question, are we talking about an arbitrary set of > =n points in the plane, or integral lattice points? The latter seems quite easy, but I haven't yet worked out how to do the former (if indeed it's true).


By David Loeffler on Wednesday, July 11, 2001 - 01:06 am:

Michael:

It was the latter actually - the set of all lattice points. But the former can be done fairly easily too. Just pick a point P not on the perpendicular bisector of AB for any of the pairs A and B of points in the set. (If we assume that there are countably many points in the set it is obvious that this may be done.) Then no two points are equidistant from P, so we can simply draw circles centred at P. As we increase the radius then the number of points inside the circle increases from 0, but it must increase in steps of 1. Result follows.

David


By Michael Doré on Wednesday, July 11, 2001 - 11:12 am:

Yes, that's pretty neat. (Actually the proof won't work for some infinite sets of points, but I was really only asking about finite sets. The result isn't true for some infinite sets which are dense in places.)

Alternatively for the lattice question, note that if a circle passes through three points with rational x,y co-ordinates, then its centre has rational x,y co-ordinates (because the centre is the intersection of the perpendicular bisectors). Now fix one point on the circumference of the circle at a point with rational co-ordinates (say (1/2,1/2)). Now shift the centre of the circle along the line y = sqrt(2)x. Note that if the circumference ever intersects two lattice points then its centre necessarily has rational co-ordinates which is impossible since the centre lies on y = sqrt(2)x. Therefore it gets one point at a time, as above.

This can be extended to the general problem by using considering the minimal field which contains the gradients of each the lines drawn from the origin to each of the points in the arrangement, but this is more complicated than your method.