error in using latin hypercube sampling 'lhsnorm'

mu =

30.2800 468.9000 5.0000 0.2042

>> cov=[21 10 40 12]

cov =

21 10 40 12

>> l=lhsnorm(mu,cov,10)

Index in position 1 exceeds array bounds (must not exceed 1).

Error in lhsnorm (line 59)

x(:,i) = norminv(x(:,i),mu(i), sqrt(sigma(i,i)));

why is it showing this error and how can i eliminate this error?

thanks in advance

Answer 1

[Walter Roberson]

When your mu is a vector of length 4, then your cov must be a 4 x 4 positive definite (covariance) matrix, not a row vector.

[MOHD UBAID]

Thank you very much Walter Roberson for you quick response .Your answer helped me a lot and i have solved the problem.