calculate distance using the below equation
3 ビュー (過去 30 日間)
古いコメントを表示
i have 2 variables, dataMatrix and queryMatrix..... dataMatrix contains 245 rows and 38 column values...... and queryMatrix contains 1 row and 38 column values...... i calculate the euclidean distance and sort the distances in ascending order as below.....
for w=1:245
dist = sum((repmat(queryMatrix(w,:),245,1)-dataMatrix).^2,2);
[sortval sortpos] = sort(dist,'ascend');
neighborIds(w,:) = sortpos(1:k);
neighborDistances(w,:) = sqrt(sortval(1:k));
end
instead of euclidean distance i want to use the distance formula in the below link.....
but i dont know to write the equation in matlab...... please can someone convert that equation to matlab..... it would be great help to me..... please do reply......
1 件のコメント
採用された回答
Matt J
2013 年 4 月 11 日
Assuming queryMatrix and dataMatrix are both 245x38, the whole thing can be done without loops,
d=permute(dataMatrix,[1,3,2]);
q=permute(queryMatrix,[3,1,2]);
num = abs(bsxfun(@minus,d,q));
den = bsxfun(@plus,abs(d),abs(q));
dist=sum(num./den,3);
[sortval sortpos] = sort(dist);
neighborIds = sortpos(1:k,:);
neighborDistances = sqrt(sortval(1:k,:));
その他の回答 (1 件)
Andrei Bobrov
2013 年 4 月 11 日
編集済み: Andrei Bobrov
2013 年 4 月 11 日
dataMatrix = randi([-10,10],10,5);
queryMatrix = randi([-10 10],1,5);
d = bsxfun(@(x,y)abs(x - y)./(abs(x) + abs(y)),...
dataMatrix,permute(queryMatrix,[3 2 1]));
d = sort(squeeze(sum(d,2)));
参考
カテゴリ
Help Center および File Exchange で Expression Analysis についてさらに検索
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!