| toxicity {robustbase} | R Documentation |
The aim of the experiment was to predict the toxicity of carboxylic acids on the basis of several molecular descriptors.
data(toxicity)
A data frame with 38 observations on the following 10 variables which are attributes for carboxylic acids:
toxicitylogKowpKaELUMOEcarbEmetRMIRTsPThe website accompanying the MMY-book: http://www.wiley.com/legacy/wileychi/robust_statistics
Maguna, F.P., N'u~nez, M.B., Okulik, N.B. and Castro, E.A. (2003) Improved QSAR analysis of the toxicity of aliphatic carboxylic acids; Russian Journal of General Chemistry 73, 1792–1798.
data(toxicity)
summary(toxicity)
plot(toxicity)
plot(toxicity ~ pKa, data = toxicity)
## robustly scale the data (to scale 1) using Qn
(scQ.tox <- sapply(toxicity, Qn))
scTox <- scale(toxicity, center = FALSE, scale = scQ.tox)
csT <- covOGK(scTox, n.iter = 2,
sigmamu = s_Qn, weight.fn = hard.rejection)
as.dist(round(cov2cor(csT$cov), 2))