plot.ru: Plot diagnostics for an ru object

Description Usage Arguments See Also Examples

View source: R/sample_diagnostics.R

Description

plot method for class "ru". For d = 1 a histogram of the simulated values is plotted with a the density function superimposed. The density is normalized crudely using the trapezium rule. For d = 2 a scatter plot of the simulated values is produced with density contours superimposed. For d > 2 pairwise plots of the simulated values are produced.

Usage

1
2
3
4
5
## S3 method for class 'ru'
plot(x, y, ..., n = ifelse(x$d == 1, 1001, 101),
  prob = c(0.1, 0.25, 0.5, 0.75, 0.95, 0.99), ru_scale = FALSE,
  rows = NULL, xlabs = NULL, ylabs = NULL, points_par = list(col =
  8))

Arguments

x

an object of class "ru", a result of a call to ru.

y

Not used.

...

Additional arguments passed on to hist, lines, contour or points.

n

A numeric scalar. Only relevant if x$d = 1 or x$d = 2. The meaning depends on the value of x$d.

  • For d = 1 : n + 1 is the number of abscissae in the trapezium method used to normalize the density.

  • For d = 2 : an n by n regular grid is used to contour the density.

prob

Numeric vector. Only relevant for d = 2. The contour lines are drawn such that the respective probabilities that the variable lies within the contour are approximately prob.

ru_scale

A logical scalar. Should we plot data and density on the scale used in the ratio-of-uniforms algorithm (TRUE) or on the original scale (FALSE)?

rows

A numeric scalar. When d > 2 this sets the number of rows of plots. If the user doesn't provide this then it is set internally.

xlabs, ylabs

Numeric vectors. When d > 2 these set the labels on the x and y axes respectively. If the user doesn't provide these then the column names of the simulated data matrix to be plotted are used.

points_par

A list of arguments to pass to points to control the appearance of points depicting the simulated values. Only relevant when d = 2.

See Also

summary.ru for summaries of the simulated values and properties of the ratio-of-uniforms algorithm.

Examples

 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
27
28
29
# Log-normal density ----------------
x <- ru(logf = dlnorm, log = TRUE, d = 1, n = 1000, lower = 0, init = 1)

## Not run: 
plot(x)

## End(Not run)

# Improve appearance using arguments to plot() and hist()
## Not run: 
plot(x, breaks = seq(0, ceiling(max(x$sim_vals)), by = 0.25),
  xlim = c(0, 10))

## End(Not run)

# Two-dimensional normal with positive association ----------------
rho <- 0.9
covmat <- matrix(c(1, rho, rho, 1), 2, 2)
log_dmvnorm <- function(x, mean = rep(0, d), sigma = diag(d)) {
  x <- matrix(x, ncol = length(x))
  d <- ncol(x)
  - 0.5 * (x - mean) %*% solve(sigma) %*% t(x - mean)
}
x <- ru(logf = log_dmvnorm, sigma = covmat, d = 2, n = 1000, init = c(0, 0))

## Not run: 
plot(x)

## End(Not run)

paulnorthrop/rust documentation built on Jan. 6, 2019, 3:09 a.m.