RndPairs {DescTools}

Create Pairs of Correlated Random Numbers

Description

Create pairs of correlated random numbers.

Usage

RndPairs(n, r, rdist1 = rnorm(n = n, mean = 0, sd = 1),
         rdist2 = rnorm(n = n, mean = 0, sd = 1), prop = NULL)

RndWord(size, length, x = LETTERS, replace = TRUE, prob = NULL)

Arguments

n

number of pairs. If length(n) > 1, the length is taken to be the number required.

r

the correlation between the two sets.

rdist1, rdist2

the distribution of the random vector X1 and X2. Default is standard normal distribution.

size

a non-negative integer giving the number of artificial words to build.

length

a non-negative integer giving the length of the words.

x

elements to choose from.

replace

Should sampling be with replacement?

prop

proportions for ordinal variable, must sum to 1.

prob

a vector of probability weights for obtaining the elements of the vector being sampled.

Value

a data.frame with 2 columns, X1 and X2 containing the random numbers

Author(s)

Andri Signorell <[e-mail]>

See Also

runif, rnorm, Random and friends

Examples

# produce 100 pairs of a normal distributed random number with a correlation of 0.7
d.frm  <- RndPairs(n=100, r=0.7)

plot(d.frm)
lines(lm(y ~ x,d.frm))

# change the distribution
d.frm  <- RndPairs(n=100, r=0.7, rdist2 = rlnorm(n = 100, meanlog = 1, sdlog = .8))
d.frm  <- RndPairs(n=100, r=0.7, rdist2 = runif(n = 100, -1, 4))

x <- StrCap(sapply(sample(3:15, 10), function(i) RndWord(1, i, x=letters)))


# produce some artificial words with defined probabilities for the letters
p <- c(6.51,1.89,3.06,5.08,17.4,1.66,3.01,4.76,7.55,0.27,1.21,3.44,2.53,
       9.78,2.51,0.79,0.02,7,7.27,6.15,4.35,0.67,1.89,0.03,0.04,1.13)
sapply(sample(3:15, 10), function(i) RndWord(1, i, x=letters, prob=p))

# produce associated ordinal variables
d.ord <- RndPairs(500, r=0.8, prop = list(c(.15, .3, .55), 
                                      c(.3, .5, .2)))
levels(d.ord$y) <- levels(d.ord$x) <- LETTERS[1:3]
PlotMosaic(table(d.ord$x, d.ord$y), las=1, main="")