MixSim (version 0.1-01)

simdataset: Dataset Simulation

Description

Simulates a datasets of sample size n given parameters of finite mixture model with Gaussian components

Usage

simdataset(n, Pi, Mu, S)

Arguments

n

sample size

Pi

vector of mixing proprtions (length K)

Mu

matrix consisting of components' mean vectors (K x p)

S

set of components' covariance matrices (p x p x K)

Value

• x
  simulated dataset (n x p)
• id
  classification vector (length n)

Details

Numbers of observations in components are assigned as a realization from multinomial distribution with probabilities given by mixing proportions

References

Maitra, R. and Melnykov, V. (200?) "Simulating data to study performance of finite mixture modeling and clustering algorithms", The Journal of Computational and Graphical Statistics.

Davies, R. (1980) "The distribution of a linear combination of chi-square random variables", Applied Statistics, 29, 323-333.

See Also

MixSim, overlap, pdplot

Examples

K <- 4
repeat{
   Q <- MixSim(BarOmega = 0.01, MaxOmega = 0.05, K = 4, p = 2)
   if (Q$fail == 0) break
}
A <- simdataset(n = 1000, Pi = Q$Pi, Mu = Q$Mu, S = Q$S)
colors <- c("red", "green", "blue", "brown")
plot(A$x, xlab = "x1", ylab = "x2", type = "n")
for (k in 1:K){
   points(A$x[A$id == k, ], col = colors[k], pch = 19, cex = 0.4)
}