s.multinom {ade4} | R Documentation |
Graph of frequency profiles (useful for instance in genetic)
Description
The main purpose of this function is to draw categories using scores and profiles by their gravity center. Confidence intervals of the average position (issued from a multinomial distribution) can be superimposed.
Usage
s.multinom(dfxy, dfrowprof, translate = FALSE, xax = 1, yax = 2,
labelcat = row.names(dfxy), clabelcat = 1, cpointcat = if (clabelcat == 0) 2 else 0,
labelrowprof = row.names(dfrowprof), clabelrowprof = 0.75,
cpointrowprof = if (clabelrowprof == 0) 2 else 0, pchrowprof = 20,
coulrowprof = grey(0.8), proba = 0.95, n.sample = apply(dfrowprof, 1, sum),
axesell = TRUE, ...)
Arguments
dfxy |
|
dfrowprof |
|
translate |
a logical value indicating whether the plot should be translated(TRUE) or not. The origin becomes the gravity center weighted by profiles. |
xax |
the column number of |
yax |
the column number of |
labelcat |
a vector of strings of characters for the labels of categories |
clabelcat |
an integer specifying the character size for the labels of categories,
used with |
cpointcat |
an integer specifying the character size for the points showing the categories,
used with |
labelrowprof |
a vector of strings of characters for the labels of profiles (rows of |
clabelrowprof |
an integer specifying the character size for the labels of profiles used with par("cex")*clabelrowprof |
cpointrowprof |
an integer specifying the character size for the points representative of the profiles used with par("cex")*cpointrowprof |
pchrowprof |
either an integer specifying a symbol or a single character to be used for the profile labels |
coulrowprof |
a vector of colors used for ellipses, possibly recycled |
proba |
a value lying between 0.500 and 0.999 to draw a confidence interval |
n.sample |
a vector containing the sample size, possibly recycled. Used |
axesell |
a logical value indicating whether the ellipse axes should be drawn |
... |
further arguments passed from the |
Value
Returns in a hidden way a list of three components :
tra |
a vector with two values giving the done original translation. |
ell |
a matrix, with 5 columns and for rows the number of profiles, giving the means,
the variances and the covariance of the profile for the used
numerical codes (column of |
call |
the matched call |
Author(s)
Daniel Chessel
Examples
par(mfrow = c(2,2))
par(mar = c(0.1,0.1,0.1,0.1))
proba <- matrix(c(0.49,0.47,0.04,0.4,0.3,0.3,0.05,0.05,0.9,0.05,0.7,0.25), ncol = 3, byrow = TRUE)
proba.df <- as.data.frame (proba)
names(proba.df) <- c("A","B","C") ; row.names(proba.df) <- c("P1","P2","P3","P4")
w.proba <- triangle.plot(proba.df, clab = 2, show = FALSE)
box()
w.tri = data.frame(x = c(-sqrt(1/2),sqrt(1/2),0), y = c(-1/sqrt(6),-1/sqrt(6),2/sqrt(6)))
L3 <- c("A","B","C")
row.names(w.tri) <- L3
s.multinom(w.tri, proba.df, n.sample = 0, coulrowprof = "black", clabelrowprof = 1.5)
s.multinom(w.tri, proba.df, n.sample = 30, coul = palette()[5])
s.multinom(w.tri, proba.df, n.sample = 60, coul = palette()[6], add.p = TRUE)
s.multinom(w.tri, proba.df, n.sample = 120, coul = grey(0.8), add.p = TRUE)
print(s.multinom(w.tri, proba.df[-3,], n.sample = 0, translate = TRUE)$tra)