Hi all,

ggplot2 (v2.0.0) does not seem to respect factor order when stacking bars in geom_bar().

##################
> dput(temp)
structure(list(phylo_sig = c(0.148740270638472, 0.148740270638472,
0.148740270638472, 0.148740270638472, 0.148740270638472), trait = c("p_corrected_percent",
"p_corrected_percent", "p_corrected_percent", "p_corrected_percent",
"p_corrected_percent"), value = c(0.031015586683513, 0.0553330412480036,
0.73674222087599, 0.114517508485538, 0.0623916427069555), varlevel = structure(c(3L,
2L, 4L, 5L, 1L), .Label = c("species", "genus", "family", "plot_code",
"Residual"), class = "factor")), .Names = c("phylo_sig", "trait",
"value", "varlevel"), row.names = c(28L, 62L, 96L, 130L, 164L
), class = "data.frame")
> str(temp)
'data.frame':    5 obs. of  4 variables:
 $ phylo_sig: num  0.149 0.149 0.149 0.149 0.149
$ trait : chr "p_corrected_percent" "p_corrected_percent" "p_corrected_percent" "p_corrected_percent" ...
 $ value    : num  0.031 0.0553 0.7367 0.1145 0.0624
 $ varlevel : Factor w/ 5 levels "species","genus",..: 3 2 4 5 1
> ggplot(temp, aes(x=trait, y = value)) + geom_bar(stat = "identity", aes(fill=varlevel, order = varlevel))
##################

The code above results in the correct legend order, but incorrect order of stacking (https://i.imgur.com/lgAxTFh.png). I have to change the order of the actual dataframe rows in order to coerce the stacking to work:

##################
> temp = temp[c(5,2,1,3,4),]
> ggplot(temp, aes(x=trait, y = value)) + geom_bar(stat = "identity", aes(fill=varlevel, order = varlevel))
##################

This results in correct stacking (https://i.imgur.com/baww43x.png). Any assistance in getting ggplot to pay attention to the factor ordering when stacking bars would be much appreciated!

Thanks,
Allie

______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Reply via email to