ggdmc (version 0.1.3.9)

dprior: Calculate Prior Probability Density for an EAM

Description

dprior is C++ function. It matches 5 string types: tnorm, beta_lu, gamma_l, lnorm_l, and constant to determine which density functions to call (via R API). For truncated normal density, dprior calls dtn_scalar, an internal Rcpp function built specific for ggdmc. Whetehr log the probability density is determined by the boolean log sent in via p.prior. This function is akin to DMC's log.prior.dmc.

Usage

dprior(pVec, pPrior)

Arguments

pVec

the user's supplied parameter vector or a sampler supplied theta/phi vector.

pPrior

a list of list usually created by prior.p.dmc to store the prior parameter setting.

Value

a double vector with probability density for each model parameter

Examples

## Use Drift-diffusion model as an example
ddm.prior <- prior.p.dmc(
dists = c("tnorm", "tnorm", "beta", "tnorm", "beta", "beta"),
p1    = c(a = 1, v = 0, z = 1, sz = 1, sv = 1, t0 = 1),
p2    = c(a = 1, v = 2, z = 1, sz = 1, sv = 1, t0 = 1),
lower = c(0,-5, NA, NA, 0, NA),
upper = c(2, 5, NA, NA, 2, NA))

view(ddm.prior)
##      mean sd lower upper log    dist  untrans
##   a     1  1     0     2   1   tnorm identity
##   v     0  2    -5     5   1   tnorm identity
##   z     1  1     0     1   1 beta_lu identity
##   sz    1  1  -Inf   Inf   1   tnorm identity
##   sv    1  1     0     2   1 beta_lu identity
##   t0    1  1     0     1   1 beta_lu identity

ddm.pVec <- c(a=1.15, v=-0.10, z=0.74, sz=1.23, sv=0.11, t0=0.87)
dprior(ddm.pVec, ddm.prior)
##         a          v          z         sz         sv         t0
##-0.5484734 -1.6008386  0.0000000 -0.9453885  0.0000000  0.0000000

## Use LBA model as an example
lba.prior <- prior.p.dmc(
dists = c("tnorm", "tnorm", "tnorm", "tnorm", "tnorm", "tnorm"),
p1    = c(A=.4, B=.6, mean_v.true=1,  mean_v.false=0,  sd_v.true=.5, t0=.3),
p2    = c(A=.1, B=.1, mean_v.true=.2, mean_v.false=.2, sd_v.true=.1, t0=.05),
lower = c(0,   0, NA, NA, 0, .1),
upper = c(NA, NA, NA, NA, NA, 1))

view(lba.prior)
##              mean   sd lower upper log  dist  untrans
## A             0.4  0.1     0   Inf   1 tnorm identity
## B             0.6  0.1     0   Inf   1 tnorm identity
## mean_v.true     1  0.2  -Inf   Inf   1 tnorm identity
## mean_v.false    0  0.2  -Inf   Inf   1 tnorm identity
## sd_v.true     0.5  0.1     0   Inf   1 tnorm identity
## t0            0.3 0.05   0.1     1   1 tnorm identity

lba.pVec <- c(A=0.398, B=0.614, mean_v.true=1.040,
              mean_v.false=-0.032, sd_v.true=0.485, t0=0.271)
dprior(lba.pVec, lba.prior)
##          A            B  mean_v.true mean_v.false    sd_v.true
##  1.3834782    1.3738466    0.6704994    0.6776994    1.3723968