How to automatically adjust the width of each facet for facet_wrap?

How to automatically adjust the width of each facet for facet_wrap?



I want to plot a boxplot using ggplot2, and i have more than one facet, each facet has different terms, as follows:


library(ggplot2)

p <- ggplot(
data=Data,
aes(x=trait,y=mean)
)

p <- p+facet_wrap(~SP,scales="free",nrow=1)

p <- p+geom_boxplot(aes(fill = Ref,
lower = mean - sd,
upper = mean + sd,
middle = mean,
ymin = min,
ymax = max,
width=c(rep(0.8/3,3),rep(0.8,9))),
lwd=0.5,
stat="identity")





as showed, the width of box in different facet is not the same, is there any way to adjust all the box at a same scale? I had tried to use facet_grid, it can automatically change the width of facets, but all facets share the same y axis.


facet_grid



Data


Data <- structure(list(SP = structure(c(3L, 3L, 3L, 4L, 4L, 4L, 4L, 4L,
4L, 4L, 4L, 4L), .Label = c("Human", "Cattle", "Horse", "Maize"
), class = "factor"), Ref = structure(c(3L, 2L, 1L, 3L, 3L, 3L,
2L, 2L, 2L, 1L, 1L, 1L), .Label = c("LMM", "Half", "Adoptive"
), class = "factor"), trait = structure(c(11L, 11L, 11L, 14L,
13L, 12L, 14L, 13L, 12L, 14L, 13L, 12L), .Label = c("cad", "ht",
"t2d", "bd", "cd", "ra", "t1d", "fpro", "mkg", "scs", "coat colour",
"ywk", "ssk", "gdd"), class = "factor"), min = c(0.324122039,
0.336486555, 0.073152049, 0.895455441, 0.849944623, 0.825248005,
0.890413591, 0.852385351, 0.826470308, 0.889139116, 0.838256672,
0.723753592), max = c(0.665536838, 0.678764774, 0.34033228, 0.919794865,
0.955018001, 0.899903826, 0.913350912, 0.957305688, 0.89843716,
0.911257005, 0.955312678, 0.817489555), mean = c(0.4919168555,
0.5360103372, 0.24320509565, 0.907436221, 0.9057516121, 0.8552899502,
0.9035394117, 0.9068819173, 0.8572309823, 0.90125638965, 0.90217769835,
0.7667208778), sd = c(0.0790133656517775, 0.09704320004497, 0.0767552215753863,
0.00611921020505611, 0.0339614482273291, 0.0199389195311925,
0.00598633573504195, 0.0332634006653858, 0.0196465508521771,
0.00592476494699222, 0.0348144156099722, 0.0271827880539459)), .Names = c("SP",
"Ref", "trait", "min", "max", "mean", "sd"), class = "data.frame", row.names = c(10L,
11L, 12L, 34L, 35L, 36L, 37L, 38L, 39L, 40L, 41L, 42L))




2 Answers
2



You can adjust facet widths after converting the ggplot object to a grob:


# create ggplot object (no need to manipulate boxplot width here.
# we'll adjust the facet width directly later)
p <- ggplot(Data,
aes(x = trait, y = mean)) +
geom_boxplot(aes(fill = Ref,
lower = mean - sd,
upper = mean + sd,
middle = mean,
ymin = min,
ymax = max),
lwd = 0.5,
stat = "identity") +
facet_wrap(~ SP, scales = "free", nrow = 1) +
scale_x_discrete(expand = c(0, 0.5)) + # change additive expansion from default 0.6 to 0.5
theme_bw()

# convert ggplot object to grob object
gp <- ggplotGrob(p)

# optional: take a look at the grob object's layout
gtable::gtable_show_layout(gp)

# get gtable columns corresponding to the facets (5 & 9, in this case)
facet.columns <- gp$layout$l[grepl("panel", gp$layout$name)]

# get the number of unique x-axis values per facet (1 & 3, in this case)
x.var <- sapply(ggplot_build(p)$layout$panel_scales_x,
function(l) length(l$range$range))

# change the relative widths of the facet columns based on
# how many unique x-axis values are in each facet
gp$widths[facet.columns] <- gp$widths[facet.columns] * x.var

# plot result
grid::grid.draw(gp)



plot comparison






Thanks so much for your help! It is the best answer for my question!

– YinLL
Oct 8 '18 at 9:41



In general, you can determine the width of a box plot in ggplot like so:


ggplot


ggplot(data= df, aes(x = `some x`, y = `some y`)) + geom_boxplot(width = `some witdth`)



In your case, you might consider setting the width of all the box plots to the range of x divided by the maximum number of elements (in the leftmost figure).






Hi! Thanks so much for your kindly help! It really works with your suggestions. I have successfully ajust all the box in the same width, but another question is that how to adjust the width of each facet? because all facets are equal width in default of facet_wrap.

– YinLL
Sep 15 '18 at 7:51






facet_wrap has the argument scales: should scales be fixed ("fixed", the default), free ("free"), or free in one dimension ("free_x", "free_y"). If you liked my answer, I would be happy if you could upvote it ;)

– Omry Atia
Sep 15 '18 at 8:01


scales






could you kindly dput your dataframe, ideally not the entire one, but use dput(head(your data frame, 10)) , so that we can use it to play with? We can help you better with it. And also the code for your plot. stackoverflow.com/questions/5963269/…

– Tjebo
Sep 15 '18 at 21:24



dput


dput(head(your data frame, 10))






i had tried the parameter "scales=free", it didn't work for facet_wrap, the width of all facets is always equal. I had made a small example at the second answer!, please have a try, Thanks in advance!

– YinLL
Sep 17 '18 at 2:04




Thanks for contributing an answer to Stack Overflow!



But avoid



To learn more, see our tips on writing great answers.



Required, but never shown



Required, but never shown




By clicking "Post Your Answer", you agree to our terms of service, privacy policy and cookie policy

Popular posts from this blog

𛂒𛀶,𛀽𛀑𛂀𛃧𛂓𛀙𛃆𛃑𛃷𛂟𛁡𛀢𛀟𛁤𛂽𛁕𛁪𛂟𛂯,𛁞𛂧𛀴𛁄𛁠𛁼𛂿𛀤 𛂘,𛁺𛂾𛃭𛃭𛃵𛀺,𛂣𛃍𛂖𛃶 𛀸𛃀𛂖𛁶𛁏𛁚 𛂢𛂞 𛁰𛂆𛀔,𛁸𛀽𛁓𛃋𛂇𛃧𛀧𛃣𛂐𛃇,𛂂𛃻𛃲𛁬𛃞𛀧𛃃𛀅 𛂭𛁠𛁡𛃇𛀷𛃓𛁥,𛁙𛁘𛁞𛃸𛁸𛃣𛁜,𛂛,𛃿,𛁯𛂘𛂌𛃛𛁱𛃌𛂈𛂇 𛁊𛃲,𛀕𛃴𛀜 𛀶𛂆𛀶𛃟𛂉𛀣,𛂐𛁞𛁾 𛁷𛂑𛁳𛂯𛀬𛃅,𛃶𛁼

Edmonton

Crossroads (UK TV series)