Last active
December 19, 2015 15:09
-
-
Save christophM/5974826 to your computer and use it in GitHub Desktop.
Multidimensional Scaling "manually"
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
## Multidimensional Scaling (MDS) | |
## the data matrix | |
X <- matrix(c(10.39, 1.81, 2.30, | |
10.31, 1.74, 2.14, | |
9.93, 1.73, 2.18, | |
10.82, 2.02, 2.70), nrow = 4, byrow = TRUE) | |
rownames(X) <- c("Argentina", "Australia", "USA", "W.Samoa") | |
colnames(X) <- c("X100m", "X800m", "marathon") | |
## center and scale | |
X <- as.matrix(adata) | |
X <- scale(X, center = TRUE, scale = TRUE) | |
X | |
## distance matrix | |
D <- as.matrix(dist(X)) | |
## squared distance | |
D2 <- D ^2 | |
D2 | |
## mean center matrix | |
n <- nrow(D2) | |
ones <- rep(1, n) | |
C <- diag(1, n) - (1 / n) * ones %*% t(ones) | |
C | |
## double center the (squared) distance matrix | |
B <- - (1 / 2) C %*% D2 %*% C | |
B | |
## find the eigen decomposition of B | |
eigen_decomposition <- eigen(B) | |
## eigenvectors | |
U <- eigen_decomposition$vectors | |
U | |
## and eigenvalues | |
G2 <- diag(eigen_decomposition$values) | |
G2 | |
## compose X with eigen values and vectors | |
Xstar <- U %*% (G2 ^ (0.5)) | |
Xstar | |
## reduce dimensions | |
k <- 2 | |
Xk <- U[, 1:k] %*% (G2 ^ 0.5)[1:k, 1:k] | |
Xk | |
plot(Xk, xlim = c(-2.5, 2.5), ylim = c(-0.7, 0.7)) | |
text(x = Xk[, 1] -0.7, y = Xk[, 2], rownames(X)) | |
## alternative: | |
cmds <- cmdscale(D) | |
cmds | |
plot(cmds) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment