Possible bug in plotIndiv()

Bugs
1

This could be a bug or maybe I am doing some basic mistake – could you have a look?

The following code generates a scores plot of a PLS-DA model to discriminate between animals according to their Time.Group

data(liver.toxicity)

X = liver.toxicity$gene

Y = as.factor(liver.toxicity$treatment[, 4])

plsda.liver <- plsda(X, Y, ncomp = 2)

plotIndiv(plsda.liver, legend=TRUE)

Rplot
Rplot805×592 18.2 KB

It looks like the legend is misleading. For example, the orange dots in the top left belong to animals 404, 405, 505, 506, 516. They are colour-coded as belonging to the “6 hours” group, but according to the data, they are all in the “18 hours” group:

liver.toxicity$treatment[,c(1,4)]

I am running mixOmics 6.12.0 on R 4.0.0

Could you check? Thanks a lot in advance

Cheers

Axel

2

Hi Axel,

Thanks for reporting this. This is in fact a serious issue which I put on top of our list now https://github.com/mixOmicsTeam/mixOmics/issues/71.

I’ll keep you updated once it is rectified.

Best,

Al

3

Hi Axel,

The issue should be fixed now and will be effective in the next release version later this week.
In the meantime you can install the GitHub version using the instructions in:

Please let us know if the problem persists, or if you face any further issues.

Best,

Al