Re: [R] MLE with parameters restricted to a defined range using bbmle

Thanks, Rolf and Ben, Both solutions worked, but I finished by contraining parameters values using optim(). Best,Bernardo Niebuhr Em Segunda-feira, 8 de Dezembro de 2014 18:36, Rolf Turner r.tur...@auckland.ac.nz escreveu: I know nothing about the bbmle package and its mle2()

[R] MLE with parameters restricted to a defined range using bbmle

Dear all, I am fitting models to data with mle2 function of the bbmle package.In specific, I want to fit a power-law distribution model, as defined here (http://arxiv.org/pdf/cond-mat/0412004v3.pdf), to data. However, one of the parameters - xmin -, must be necessarily greater than zero. What

Re: [R] MLE with parameters restricted to a defined range using bbmle

I know nothing about the bbmle package and its mle2() function, but it is a general truth that if you need to constrain a parameter to be positive in an optimisation procedure a simple and effective approach is to reparameterize using exp(). I.e. represent xmin as exp(lxmin) (say) and use

Re: [R] MLE with parameters restricted to a defined range using bbmle

Rolf Turner r.turner at auckland.ac.nz writes: I know nothing about the bbmle package and its mle2() function, but it is a general truth that if you need to constrain a parameter to be positive in an optimisation procedure a simple and effective approach is to reparameterize using