- attached plotIndiv.R
plotIndiv
I renamed plot.cent as plot.centroid
- should plot.star only applied when argument group is provided? OK
- should plot.centroid only applied when argument group is provided? OK
- add examples with plot.star and plot.centroid in running test OK, partly
- update help file OK
- Indicate the centroid as a thick black 'star' otherwise it does not show on its own
Start from the 3 files named 'issue plotIndiv in devel.graphique'
Comments (8)
-
-
reporter - edited description
-
reporter plotIndiv
Code: function(object, comp = NULL, rep.space = NULL, blocks = NULL, ind.names = TRUE, group, col.per.group, style = "lattice", plot.ellipse = TRUE, ellipse.level = 0.95, plot.centroid = TRUE, plot.star = FALSE, main = NULL, add.legend = FALSE, X.label = NULL, Y.label = NULL, Z.label = NULL, abline.line = FALSE, xlim = NULL, ylim = NULL, col, cex, pch, alpha = 0.2, axes.box = "box")
Docs: function(object, comp = NULL, rep.space = NULL, blocks = NULL, ind.names = TRUE, group, col.per.group, style = "ggplot2", plot.ellipse = FALSE, ellipse.level = 0.95, plot.centroid = FALSE, plot.star = FALSE, main = NULL, add.legend = FALSE, X.label = NULL, Y.label = NULL, Z.label = NULL, abline.line = FALSE, xlim = NULL, ylim = NULL, col, cex, pch, alpha = 0.2, axes.box = "box") Mismatches in argument default values: Name: 'style' Code: "lattice" Docs: "ggplot2" Name: 'plot.ellipse' Code: TRUE Docs: FALSE Name: 'plot.centroid' Code: TRUE Docs: FALSE
-
reporter plot of individuals for objects of class 'plsda' or 'splsda'
----------------------------------------------------
data(breast.tumors) X <- breast.tumors$gene.exp Y <- breast.tumors$sample$treatment
splsda.breast <- splsda(X, Y,keepX=c(10,10),ncomp=2)
only shows 1 star when you should see 2 (by default from group)
plotIndiv(splsda.breast, plot.centroid = T)
also, is there a way to color the stars according to the group (internal) or col.group?
-
reporter Hello, I havent tested your push from last night (conflicts) but I have further issues.
Before I was working with the updated plotIndiv script (see attached) and it was giving the right figure - except! that the legend key was wrong.
Now, with the non-pushed mixOmics devel.graphique I get this (see attached) * title block X still there for a sPLSDA * legend lty does not work well. Could we put pch = 16 by default please? * legend key is right
-
reporter - attached plotIndiv_updated.R
-
reporter Ok, forget about the above I think I had the wrong version.
Testing the plotIndiv you pushed, we have issues with the legend and labelling. See attached. The running script is on bitbucket (noted #FB). Seems to be an issue when there are more than 2 classes. Test thoroughly please.
-
reporter - changed status to resolved
V5.3
- Log in to comment