Description Usage Arguments Value Author(s) References Examples
A function to obtain the optimal Bayesian sample size via a decision-theoretic approach for estimating the mean of the Birbaum-Saunders distribution.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | bss.dt.bs(
loss = "L1",
a1 = 8,
b1 = 50,
a2 = 8,
b2 = 50,
cost = 0.01,
rho = 0.05,
gam = 1,
l = 1,
nmin = 2,
nmax = 2000,
nlag = 200,
nrep = 6L,
lrep = 100,
npost = 500,
nburn = 500,
thin = 20L,
scale = 1L,
plots = TRUE,
prints = TRUE,
save.plot = FALSE,
diagnostic = FALSE,
mc.preschedule = FALSE,
...
)
|
loss |
L1 (Absolute loss), L2 (Quadratic loss), L3 (Weighted loss) and L4 (Half loss) representing the loss function used. The default is absolute loss function. |
a1 |
hyperparameter of the prior distribution for beta. The default is 3. |
b1 |
hyperparameter of the prior distribution for beta. The default is 2. |
a2 |
hyperparameter of the prior distribution for |
b2 |
hyperparameter of the prior distribution for |
cost |
a positive real number representing the cost of colect one observation. The default is 0.010. |
rho |
a number in (0, 1). The probability of the credible interval is 1-rho. Only for loss function L3. The default is 0.95. |
gam |
a positive real number connected with the credible interval when using loss function L4. The default is 0.5. |
l |
xxx |
nmin |
a positive integer representing the minimum number for compute the Bayes risk. Default is 2. |
nmax |
a positive integer representing the maximum number for compute the Bayes risk. Default is 100. |
nlag |
a positive integer representing the lag in the n's used to compute the Bayes risk. Default is 10. |
nrep |
a positive integer representing the number of samples taken for each n. |
lrep |
a positive integer representing the number of samples taken for S_n. Default is 100. |
npost |
a positive integer representing the number of values to draw from the posterior distribution of the mean. Default is 100. |
nburn |
a positive constant for the sampling method. |
thin |
a positive constant for the sampling method. |
scale |
a positive constant for the sampling method. |
plots |
Boolean. If TRUE (default) it plot the estimated Bayes risks and the fitted curve. |
prints |
Boolean. If FALSE (default) the output is a list. |
save.plot |
Boolean. If TRUE, the plot is saved to an external file. The default is FALSE. |
diagnostic |
xxx. |
mc.preschedule |
xxx |
... |
Currently ignored. |
An integer representing the optimal sample size.
Eliardo G. Costa eliardocosta@ccet.ufrn.br and Manoel Santos-Neto manoel.ferreira@ufcg.edu.br
Costa, E.G., Paulino, C.D., and Singer, J. M. (2019). Sample size determination to evaluate ballast water standards: a decision-theoretic approach. Tech. rept. University of Sao Paulo.
1 2 |
#bss.dt.bs(loss="L1", plot=TRUE, lrep=10, npost=10)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.