ddesire {desire} | R Documentation |
Generic density, distribution, quantile and random number generation functions for desirability functions.
## Default S3 method: ddesire(x, f, mean = 0, sd = 1) ## Default S3 method: pdesire(q, f, mean = 0, sd = 1) ## Default S3 method: qdesire(p, f, mean = 0, sd = 1) ## Default S3 method: rdesire(n, f, mean = 0, sd = 1) ## Default S3 method: edesire(f, mean, sd) ## Default S3 method: vdesire(f, mean, sd)
x,q |
Vector of quantiles. |
p |
vector of probabilies. |
n |
number of observations. |
f |
desirability function |
mean |
vector of means. |
sd |
vector of standard deviations. |
'ddesire' gives the density, 'pdesire' gives the distribution
function, 'qdesire' gives the quantile function, and 'rdesire'
generates random deviates.
'edesire' and 'vdesire' return the expectation and variance of the
function.
The default implementations for pdesire
, qdesire
,
edesire
and vdesire
are only approximations obtained by
estimating the desired property from a random sample.
Heike Trautmann trautmann@statistik.tu-dortmund.de, Detlef Steuer steuer@hsu-hamburg.de and Olaf Mersmann olafm@statistik.tu-dortmund.de
For desirability functions:
harrington1
and
harrington2
data(Chocolate) ## Fit linear model to data: m.d90 <- lm(d90 ~ rt + as + I(rt^2) + I(as^2) + rt:as, Chocolate) m.Fe <- lm(Fe ~ rt + as + I(rt^2) + I(as^2) + rt:as, Chocolate) ## Define desirability functions: d.d90 <- harrington2(21, 22, 1) d.Fe <- harrington1(22, 0.8, 28, 0.2) ## Plot density of desirability in rt=30, as=50: df <- data.frame(rt=30, as=50) y.Fe <- predict(m.Fe, df) sigma.Fe <- summary(m.Fe)$sigma y.d90 <- predict(m.d90, df) sigma.d90 <- summary(m.d90)$sigma ## Plot curve of density function: opar <- par(mfrow=c(2,1)) curve(ddesire(x, d.d90, y.d90, sigma.d90), 0, 1, main="d.90", n=202) curve(ddesire(x, d.Fe, y.Fe, sigma.Fe), 0, 1, main="Fe", n=202) par(opar) ## Integrate: integrate(function(x) ddesire(x, d.d90, y.d90, sigma.d90), 0, 1) integrate(function(x) ddesire(x, d.Fe, y.Fe, sigma.Fe), 0, 1)