pdist2

Pairwise distances between two sets of observations.

📝Syntax

D = pdist2(X, Y)
D = pdist2(X, Y, distance)
D = pdist2(X, Y, distance, distanceParameter)
D = pdist2(..., 'Smallest', K)
D = pdist2(..., 'Largest', K)
[D, I] = pdist2(..., 'Smallest', K)
[D, I] = pdist2(..., 'Largest', K)

📥Input Arguments

Parameter	Description
X	real full numeric matrix. Rows are observations.
Y	real full numeric matrix with the same number of columns as X.
distance	distance name or function handle. Supported names include euclidean, squaredeuclidean, sqeuclidean, cityblock, chebychev, chebyshev, cosine, correlation, hamming, jaccard, minkowski, seuclidean, mahalanobis, spearman, fasteuclidean, and fastseuclidean.
distanceParameter	optional parameter for minkowski, seuclidean, or mahalanobis.
K	positive integer number of smallest or largest distances to return for each row of Y.

📤Output Arguments

Parameter	Description
D	distance matrix. Without Smallest or Largest, D has size size(X, 1)-by-size(Y, 1). With Smallest or Largest, D has size min(K, size(X, 1))-by-size(Y, 1).
I	indices of rows in X for the selected distances. I is available only with Smallest or Largest.

📄Description

pdist2 computes pairwise distances between rows of X and rows of Y. Built-in distances return NaN when either row contains NaN. A function handle distance must accept one row of X and all rows of Y, and return one distance per row of Y.

💡Examples

Compute pairwise Euclidean distances.

X = [0 0; 1 0];
Y = [0 0; 0 2];
D = pdist2(X, Y)

Compare several distance metrics.

X = [1 0; 0 1];
Y = [1 0; 1 1];
Dcos = pdist2(X, Y, 'cosine')
Dhamming = pdist2(X, Y, 'hamming')

Find selected distances and row indices.

X = [0 0; 1 0; 0 3];
Y = [0 0; 0 2];
[D, I] = pdist2(X, Y, 'euclidean', 'Smallest', 2)

Used Functions

kmeans kmedoids silhouette

Edit this page on GitHub