ggdmc (version 0.2.5.2)

rplba0: Piecewise LBA model

Description

Density and random generation of the PLBA Model Type 0, 1, and 2.

Usage

rplba0(n, A, b, t0, mean_v, mean_w, sd_v, rD, swt, ncore = 1L,
  debug = FALSE)
rplba1(n, A, b, t0, mean_v, mean_w, sd_v, rD, swt, ncore = 1L,
  debug = FALSE)
rplba2(n, A, b, t0, mean_v, mean_w, sd_v, sd_w, rD, swt, ncore = 1L,
  debug = FALSE)
rplba3(n, A, B, C, mean_v, mean_w, sd_v, sd_w, rD, tD, swt, t0)
n1PDF_plba1(x, A, b, mean_v, sd_v, t0, mean_w, rD, swt, n, h, ncore, debug)
n1PDF_plba2(x, A, b, mean_v, sd_v, t0, mean_w, sd_w, rD, swt, n, h, ncore,
  debug)
n1PDF_plba3(x, n, A, B, C, mean_v, sd_v, mean_w, sd_w, rD, tD, swt, t0, h)
rplba1R(n, A, b, t0, mean_v, mean_w, sd_v, rD, swt)
rplba2R(n, A, b, t0, mean_v, mean_w, sd_v, sd_w, rD, swt)
rplba3R(n = 10, pVec = c(A1 = 1.5, A2 = 1.5, B1 = 1.2, B2 = 1.2, C1 =
  0.3, C2 = 0.3, v1 = 3.32, v2 = 2.24, w1 = 1.51, w2 = 3.69, sv1 = 1, sv2 =
  1, sw1 = 1, sw2 = 1, rD = 0.3, tD = 0.3, swt = 0.5, t0 = 0.08))

Arguments

n

number of observations.

A

upper bound of start point. It can be an integer or a 2-element vector.

b

response threshold. It can be an integer or a 2-element vector.

t0

nondecision time (in second)

mean_v

stage 1 mean drift rate. It must be a 2-element vector

mean_w

stage 2 mean drift rate. It must be a 2-element vector

sd_v

common standard deviation of the piece 1 drift rates. If sd_w is not present, this will also be used as the piece 2 drift rate standard deviation, which cannot be negative.

rD

drift rate delay (in second)

swt

switch time (in second)

ncore

number of CPU cores for running Open MP.

debug

internal debug switch

sd_w

standard deviation of the piece 2 drift rates

B

first stage traveling distance

C

second stage traveling distance

tD

threshold delay time

x

vector of quantiles.

h

bandwidth for the kernel function

pVec

PLBA parameter vector

Value

a [RT R] matrix (C++) or a data frame (R)

References

Holmes, R. W., Trueblood, J. & Heathcote, A. (2016). A new framework for modeling decisions about changing information: The Piecewise Linear Ballistic Accumulator model. Cognitive Psychology, 85, 1--29, doi: http://dx.doi.org/10.1016/j.cogpsych.2015.11.002.Approximate

Examples

# NOT RUN {
#########################################################################80
## rplba1
#########################################################################80
# }
# NOT RUN {
n <- 2^20; n
A <- 1.5
b <- 2.7
mean_v <- c(3.3, 2.2)
mean_w <- c(1.5, 3.7)
sd_v <- c(1, 1)
rD    <- .3
swt   <- .5
t0    <- .08
ncore <- 12
dat1 <- rplba1R(n, A, b, t0, mean_v, mean_w, sd_v, rD, swt)
dat2 <- rplba1(n, A, b, t0, mean_v, mean_w, sd_v, rD, swt, ncore)
dat3 <- ppda::rplba1(n, A, b, t0, mean_v, mean_w, sd_v, rD, swt)

dat1r1 <- dat1[dat1[, 2] == 1, 1]
dat1r2 <- dat1[dat1[, 2] == 2, 1]
dat2r1 <- dat2[dat2[, 2] == 1, 1]
dat2r2 <- dat2[dat2[, 2] == 2, 1]
dat3r1 <- dat3[dat3[, 2] == 1, 1]
dat3r2 <- dat3[dat3[, 2] == 2, 1]

xlim <- c(0, 3)
## Check if two methods produce SPDF overlaping with each other
par(mfrow = c(4, 2), mar = c(4, 5.3, 0.82, 1))
hist(dat1r1, breaks = "fd", freq = FALSE, main = "Choice1 R", xlim = xlim)
hist(dat1r2, breaks = "fd", freq = FALSE, main = "Choice2 R", xlim = xlim)
hist(dat2r1, breaks = "fd", freq = FALSE, main = "Choice1 C++", xlim = xlim)
hist(dat2r2, breaks = "fd", freq = FALSE, main = "Choice2 C++", xlim = xlim)
hist(dat3r1, breaks = "fd", freq = FALSE, main = "Choice1 GPU", xlim = xlim)
hist(dat3r2, breaks = "fd", freq = FALSE, main = "Choice2 GPU", xlim = xlim)

par(mfrow = c(1, 2))
hist(dat1r1, breaks = "fd", freq = FALSE, main = "Choice1 R, C++, & GPU",
  xlim = xlim, ylim = c(0, 3))
hist(dat2r1, breaks = "fd", freq = FALSE, add = TRUE, col = "lightblue")
hist(dat3r1, breaks = "fd", freq = FALSE, add = TRUE, col = "lightgreen")

hist(dat1r2, breaks = "fd", freq = FALSE, main = "Choice2 R, C++ & GPU",
  xlim = xlim, ylim = c(0, 3))
hist(dat2r2, breaks = "fd", freq = FALSE, add = TRUE, col = "lightblue")
hist(dat3r2, breaks = "fd", freq = FALSE, add = TRUE, col = "lightgreen")
# }
# NOT RUN {
#############20
## rplba2    ##
#############20
# }
# NOT RUN {
n <- 2^15
ncore <- 4
A <- c(1.5, 1.5)
b <- c(2.7, 2.7)
mean_v <- c(3.3, 2.2)
mean_w <- c(1.5, 3.7)
sd_v <- c(1, 1)
sd_w <- c(1, 1)
rD <- .3
swt <- .5
t0 <- .08
dat1 <- rplba2R(n, A, b, t0, mean_v, mean_w, sd_v, sd_w, rD, swt)
dat2 <- rplba2(n, A, b, t0, mean_v, mean_w, sd_v, sd_w, rD, swt, ncore)
dat3 <- rplba2(n, A, b, t0, mean_v, mean_w, sd_v, sd_w, rD, swt)
dat4 <- rplba2_test(n, A, b, t0, mean_v, mean_w, sd_v, sd_w, rD, swt)

dat1r1 <- dat1[dat1[, 2] == 1, 1]
dat1r2 <- dat1[dat1[, 2] == 2, 1]
dat2r1 <- dat2[dat2[, 2] == 1, 1]
dat2r2 <- dat2[dat2[, 2] == 2, 1]
dat3r1 <- dat3[dat3[, 2] == 1, 1]
dat3r2 <- dat3[dat3[, 2] == 2, 1]
dat4r1 <- dat4[dat4[, 2] == 1, 1]
dat4r2 <- dat4[dat4[, 2] == 2, 1]

wesanderson::wes_palette("Royal1")
palettes  <- wesanderson::wes_palettes$GrandBudapest
palettes2 <- wesanderson::wes_palettes$GrandBudapest2
xlim <- c(0, 3)
## Check if two methods produce SPDF overlaping with each other
par(mfrow = c(4, 2), mar = c(4, 5.3, 0.82, 1))
hist(dat1r1, breaks = "fd", freq = FALSE, main = "Choice1 R", xlim = xlim)
hist(dat1r2, breaks = "fd", freq = FALSE, main = "Choice2 R", xlim = xlim)
hist(dat2r1, breaks = "fd", freq = FALSE, main = "Choice1 C++", xlim = xlim)
hist(dat2r2, breaks = "fd", freq = FALSE, main = "Choice2 C++", xlim = xlim)
hist(dat3r1, breaks = "fd", freq = FALSE, main = "Choice1 GPU", xlim = xlim)
hist(dat3r2, breaks = "fd", freq = FALSE, main = "Choice2 GPU", xlim = xlim)
hist(dat4r1, breaks = "fd", freq = FALSE, main = "Choice1 test", xlim = xlim)
hist(dat4r2, breaks = "fd", freq = FALSE, main = "Choice2 test", xlim = xlim)

par(mfrow = c(1, 2))
hist(dat1r1, breaks = "fd", freq = FALSE, main = "Choice1 R, C++, & GPU",
  xlim = xlim, ylim = c(0, 3))
hist(dat2r1, breaks = "fd", freq = FALSE, add = TRUE, col = palettes[1])
hist(dat3r1, breaks = "fd", freq = FALSE, add = TRUE, col = palettes[2])
hist(dat4r1, breaks = "fd", freq = FALSE, add = TRUE, col = palettes[4])

hist(dat1r2, breaks = "fd", freq = FALSE, main = "Choice2 R, C++ & GPU",
  xlim = xlim, ylim = c(0, 3))
hist(dat2r2, breaks = "fd", freq = FALSE, add = TRUE, col = palettes2[1])
hist(dat3r2, breaks = "fd", freq = FALSE, add = TRUE, col = palettes2[2])
hist(dat4r2, breaks = "fd", freq = FALSE, add = TRUE, col = palettes2[3])


# }