post_RA_fits {ra4bayesmeta} | R Documentation |
Computes a table of Hellinger distances between marginal posterior distributions
for different parameters in the NNHM
between the actual model fits in fits.actual
and
the benchmark fits in fits.bm
.
All fits should be based on the same data set.
post_RA_fits(fits.actual, fits.bm)
fits.actual |
a list of model fits of class bayesmeta, computed with
the |
fits.bm |
a list of model fits of class bayesmeta, computed with
the |
Suggestions for posterior benchmarks are provided
in Ott et al. (2020, Sections 2.5 and 2.6) and they
can be computed using the function fit_models_RA
.
A matrix of Hellinger distance estimates between marginal posteriors
with n.bm columns and n.act*(k+3) rows,
where n.bm=length(fits.bm
) is the number of benchmark fits specified,
n.act=length(fits.actual
) the number of actual fits specified
and
k the number of studies in the meta-analysis data set
(so that there are k+3 parameters Ψ \in \{ μ, τ, θ_1, ..., θ_k, θ_{new} \} of potential interest in the NNHM).
The columns of the matrix give the following Hellinger distance estimates between two marginal posteriors (for the parameter of interest Ψ varying with rows) induced by the following two heterogeneity priors (from left to right):
H(po_bm_1, po_act) |
first benchmark prior bm_1 inducing the fit |
H(po_bm_2, po_act) |
second benchmark prior bm_2 inducing the fit |
... |
... |
H(po_bm_n.bm, po_act) |
last benchmark prior bm_n.bm inducing the fit |
The actual heterogenity prior and the parameter of interest Ψ vary with the rows in the following order:
mu, pri_act_1 |
Ψ=μ and first actual prior in |
mu, pri_act_2 |
Ψ=μ and second actual prior in |
... |
... |
mu, pri_act_n |
Ψ=μ and nth actual prior in |
tau, pri_act_1 |
Ψ=τ and first actual prior in |
... |
... |
tau, pri_act_n |
Ψ=τ and nth actual prior |
theta_1, pri_act_1 |
Ψ=θ_1 and first actual prior |
... |
... |
theta_k, pri_act_n |
Ψ=θ_k and nth actual prior |
theta_new, pri_act_1 |
Ψ=θ_{new} and first actual prior |
... |
... |
theta_new, pri_act_n |
Ψ=θ_{new} and nth actual prior |
Ott, M., Plummer, M., Roos, M. How vague is vague? How informative is informative? Reference analysis for Bayesian meta-analysis. Manuscript submitted to Statistics in Medicine. 2020.
bayesmeta
in the package bayesmeta,
fit_models_RA
,
post_RA
,
pri_RA_fits
# for aurigular acupuncture (AA) data set data(aa) # compute the model fits # actual standard half-normal and half-Cauchy heterogeneity priors fits <- fit_models_RA(df=aa, tau.prior= list(function(t)dhalfnormal(t, scale=1), function(t)dhalfcauchy(t, scale=1)))[[1]] # benchmark fits under SGC(m_inf), Jeffreys' and SIGC(M_inf) priors fits.bm.post <- fits[c(1,5,4)] fits.actual <- fits[c(6,7)] post_RA_fits(fits.actual=fits.actual, fits.bm=fits.bm.post)