mle_gamma_lnorm {dvmisc} | R Documentation |
Each observation is assumed to be the product of a Gamma(alpha, beta) and
Lognormal(mu, sigsq) random variable. Performs maximization via
nlminb
. alpha and beta correspond to the shape and scale
(not shape and rate) parameters described in GammaDist
,
and mu and sigsq correspond to meanlog and sdlog^2 in
Lognormal
.
mle_gamma_lnorm(x, gamma_mean1 = FALSE, lnorm_mean1 = TRUE,
integrate_tol = 1e-08, estimate_var = FALSE, ...)
x |
Numeric vector. |
gamma_mean1 |
Whether to use restriction that the Gamma variable is mean-1. |
lnorm_mean1 |
Whether to use restriction that the lognormal variable is mean-1. |
integrate_tol |
Numeric value specifying the |
estimate_var |
Logical value for whether to return Hessian-based variance-covariance matrix. |
... |
Additional arguments to pass to |
List containing:
Numeric vector of parameter estimates.
Variance-covariance matrix (if estimate_var = TRUE
).
Returned nlminb
object from maximizing the
log-likelihood function.
Akaike information criterion (AIC).
# Generate 1,000 values from Gamma(0.5, 1) x Lognormal(-1.5/2, 1.5) and
# estimate parameters
## Not run:
set.seed(123)
x <- rgamma(1000, 0.5, 1) * rlnorm(1000, -1.5/2, sqrt(1.5))
mle_gamma_lnorm(x, control = list(trace = 1))
## End(Not run)