finding the closest grid

2 views (last 30 days)
Matt Learner
Matt Learner on 16 Feb 2012
Edited: Daniel on 15 Oct 2013
I have 2 variables and each of 3 discretization say for example, first variable have values (2,4,6) and second variable have (1,3,5). so out of these I can make a 9 combination which looks something like this: grid = {(2,1), (2,3), (2,5), (4,1), (4,3), (4,5), (6,1), (6,3), (6,5)}
Suppose if I have a combination (2.8, 4.2) and I would like to find the closest value out of my 9 combination for it
How can I do that ?
Thanks

Accepted Answer

Walter Roberson
Walter Roberson on 16 Feb 2012
nearestx = interp1([2,4,6], [2,4,6], comb(:,1), 'nearest');
nearesty = interp1([1,3,5], [1,3,5], comb(:,2), 'nearest');
There are other ways as well.
  2 Comments
Matt Learner
Matt Learner on 17 Feb 2012
Please can you let me know what are the other ways. Cos, I have around 64 combinations obtained from 2 variables and I have to select the nearest combination (out of 64) for the given combination
Walter Roberson
Walter Roberson on 17 Feb 2012
The above will handle that situation as well.
var1vals = sort([2, 4, 6, 7, -3, 11, pi^2, sqrt(30)]);
var2vals = sort([1, 3, 5, 9, 1.15, 33, -5, -8]);
nearestx = interp1(var1vals, var1vals, comb(:,1), 'nearest');
nearesty = interp1(var2vals, var2vals, comb(:,2), 'nearest');
An alternative:
var1vals = sort([2, 4, 6, 7, -3, 11, pi^2, sqrt(30)]);
var2vals = sort([1, 3, 5, 9, 1.15, 33, -5, -8]);
[mindiffx, minidxx] = min(bsxfun(@(a,b) abs(a-b), comb(:,1), var1vals), [], 2);
[mindiffy, minidxy] = min(bsxfun(@(a,b) abs(a-b), comb(:,2), var2vals), [], 2);
nearestx = var1vals(minidxx);
nearesty = var2vals(minidxy);
Note that both of these approaches handle all of the combinations at the same time, if the combinations are stored in the matrix "comb" with the first column being the first variable and the second column being the second variable.

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!