Hello I want to estimate the covariance matrix of the likelihood f(x1,x2,x3)=f(x2|x1)f(x3|x2)f(x1), where f(x2|x1) follows a Binomial distribution with parameters (2, 0.2), f(x3|x2) follows a Binomial distribution with parameters (2, 0.8) and f(x1) follows a Binomial distribution with parameters (2, 0.1). Could you please suggest a way of doing it using log likelihood and optimize? Many thanks Tonia Marks [[alternative HTML version deleted]]