Hello everyone. I am using the rstan to do the fitting, my code is below
stan.fit = stan(file = 'DSP with WL.stan',
data = list(n = N,
cmatrix = C_Matrix,
y = Y),
algorithm = "HMC",
control = list(adapt_delta = 0.8,
max_treedepth = 10),
chains = 1,iter = staniter,warmup = stanwarmup,refresh = 1);
In my code, I set algorithm = “HMC”. As I know in the HMC algorithm, the gradient of the log posterior is need, but it didn’t ask me to give the gradient of the log posterior. I want to know did I do it in the wrong way or stan will calculate the gradient automatically?