Customed distance function Haversine
16 views (last 30 days)
Show older comments
Miguel Angel Sanchez Aportela
on 9 Jun 2021
Commented: Scott MacKenzie
on 13 Jul 2021
For a proyect I want to use pdist2 (Pairwise distance between two sets of observations) but I need an specific function.
I have two data sets of different sizes, one of which is a nx2 matrix of latitude, longitude.
I'm trying to calculate Haversine distance but I don't know how to apply the funtion in this case. Would you help correcting my function? Thank you in advance.
axa = pdist2(latlon1(:, 1:2), latlon2(:, 1:2), @haversine);
function [dist] = haversine(lat1,lon1,lat2,lon2)
dist = 2 * 6372.8 * asin(sqrt(sind((lat2-lat1)/2)^2 + cosd(lat1) * cosd(lat2) * sind((lon2 - lon1)/2)^2));
end
1 Comment
SALAH ALRABEEI
on 21 Jun 2021
pdist2 has only three metric distances ( 'seuclidean', 'minkowski', or 'mahalanobis').
So since your lon and lat are not of the same length; you can create grid
[x,y]=meshgrid(lon,lat);
Accepted Answer
Scott MacKenzie
on 21 Jun 2021
Edited: Scott MacKenzie
on 26 Jun 2021
You need to re-work your haversine function, as per the requirements for distance functions used with pdist2. A custom distance function for pdist2 needs two input arguments and the first must specify just a single observation:
axa = pdist2(latlon1(1, 1:2), latlon2(:, 1:2), @haversine); % repeat for other values in latlon1
function [dist] = haversine(ZI, ZJ)
lat1 = ZI(1,1);
lon1 = ZI(1,2);
lat2 = ZJ(:,1);
lon2 = ZJ(:,2);
dist = 2 * 6372.8 * asin(sqrt(sind((lat2-lat1)/2)^2 + cosd(lat1) * cosd(lat2) * sind((lon2 - lon1)/2)^2));
end
2 Comments
Scott MacKenzie
on 13 Jul 2021
@Miguel Angel Sanchez Aportela You're welcome. Glad to help. Good luck with your research.
More Answers (0)
See Also
Categories
Find more on Statistics and Machine Learning Toolbox in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!