propagate (version 1.3-0)

plot.propagate: Plotting function for 'propagate' objects

Description

Creates a histogram (blue) of the Monte Carlo output, along with a density curve (red), the \(100(1 - \alpha)\) % coverage intervals (GUM uncertainty framework (GUF): dashed black; Monte Carlo, probabilistically symmetric: solid black), median (orange) and mean (green) of the Monte Carlo output, with \(\alpha\) = alpha of the propagate call.

Usage

# S3 method for propagate
plot(x, logx = FALSE, breaks = 100, ...)

Value

A plot as described above.

Arguments

x

an object returned from propagate.

logx

logical. If TRUE, the histogram and density curve show the log-transformed Monte Carlo output \(\log(y)\) on a linear abscissa (labelled "log(y)"), and the vertical lines are drawn at the logarithms of the interval limits, median and mean. Non-positive values give non-finite logarithms and are not shown.

breaks

numeric, passed to hist: a suggestion for the number of bins of the histogram.

...

other parameters to hist.

Author

Andrej-Nikolai Spiess

Details

To prevent a few extreme values from compressing the histogram, the Monte Carlo output is restricted to the values between its 0.1 % and 99.9 % quantiles, i.e. the 0.1 % most extreme values on each side are omitted from the histogram and density curve (not from the intervals, median and mean, which are those of the propagate results).

Examples

EXPR1 <- expression(x^2 * sin(y))
x <- c(5, 0.01)
y <- c(1, 0.01)
DF1 <- cbind(x, y)
RES1 <- propagate(expr = EXPR1, data = DF1, type = "stat", 
                  nsim = 100000, alpha = 0.01, check = FALSE)
plot(RES1)