Graph to Analyse the Relation between a Score and Quantitative Variables
represents the graphs to analyse the relation between a score and quantitative variables.
sco.quant (score, df, fac = NULL, clabel = 1, abline = FALSE, sub = names(df), csub = 2, possub = "topleft")
score |
a numeric vector |
df |
a data frame which rows equal to the score length |
fac |
a factor with the same length than the score |
clabel |
character size for the class labels (if any) used with |
abline |
a logical value indicating whether a regression line should be added |
sub |
a vector of strings of characters for the labels of variables |
csub |
a character size for the legend, used with |
possub |
a string of characters indicating the sub-title position ("topleft", "topright", "bottomleft", "bottomright") |
Daniel Chessel
w <- runif(100, -5, 10) fw <- cut (w, 5) levels(fw) <- LETTERS[1:5] wX <- data.frame(matrix(w + rnorm(900, sd = (1:900) / 100), 100, 9)) sco.quant(w, wX, fac = fw, abline = TRUE, clab = 2, csub = 3)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.