Drawing samples from MulitnormalDistribution

Question

I have a question concerning generating samples for a given mean vector and covariance matrix.

mu = {0.422084, 0.518201, -1.13857, -1.34951, 0, 2};

cov = {{0.000749594187095, 0.000670223129197, -0.006164273736522, 
  -0.004243056424882, 0.011372890644823, 0.001311870031246}, 
 {0.000670223129197, 0.001244975293461, -0.004710830597348, 
  -0.010103508140104, 0.004972780726833, 0.009465313700532}, 
 {-0.006164273736522, -0.004710830597348, 0.094116175041483, 
  0.042219205628345, -0.371146407307241, -0.019976547027344}, 
 {-0.004243056424882, -0.010103508140104, 0.042219205628345, 
  0.16237356863464, -0.072506782444349, -0.393141309425425}, 
 {0.011372890644823, 0.004972780726833, -0.371146407307241, 
  -0.072506782444349, 3.22993771179459, 0.261221262531002}, 
 {0.001311870031246, 0.009465313700532, -0.019976547027344, 
  -0.393141309425425, 0.261221262531002, 2.6882275522931}}

dist= 
  MultinormalDistribution[mu, cov];

I want to draw some samples between -1 and 1 but i cant even generate one sample with:

Anyone knows why this is the case? Am I missing something? RandomVariate[TruncatedDistribution[{-1, 1}, dist], 1]

Answer 1

The first argument of TruncatedDistribution should be list of pairs with length equal to the dimension of dist:

RandomVariate[TruncatedDistribution[ConstantArray[{-1, 1}, Length @ mu], dist], 3]