How do I erase these labels in ggbarstats plot?

Question

This is my code:

daten_short<-structure(list(Asthma = c(0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 
                                       0, 0, 0, 1, 0, 0, 1, 1), Genetische_PräAs = c("Nein", "Nein", 
                                                                                     "Nein", "Nein", "Ja", "Nein", "Ja", "Nein", "Ja", "Nein", "Nein", 
                                                                                     "Nein", "Nein", "Nein", "Nein", "Nein", "Ja", "Nein", "Nein", 
                                                                                     "Ja")), row.names = c(NA, -20L), class = c("tbl_df", "tbl", "data.frame"
                                                                                     ))
library(ggstatsplot)
ggbarstats(
 data=daten_short,
 x= Asthma,
 y= Genetische_PräAs,
label="both",
digits=3)

This is the output of the code: bar plots

I only want to see the p-value and the number of observed values. I want to erase the rest(x^2 Person, Cramers V, 95%CI).

How do I do it?

Did you read the documentation (help(ggstatsplot::ggbarstats))? — Friede
– Friede, Commented Nov 25 at 12:33
Please post a minimal reproducible example for us to reproduce and improve your results. You can strip down the data to the bare minimum for that. — André
– André, Commented Nov 25 at 14:04
I have edited it with sample data. The code is reproducible now. — Sourish Rakshit
– Sourish Rakshit, Commented Nov 25 at 14:24

Friede · Accepted Answer · 2025-11-26 20:23:41Z

2

I only want to see the p-value and the number of observed values.

If we are certain there isn't any direct possibility via an argument of ggstatsplot::ggbarstats() (haven't checked) try

ggstatsplot::ggbarstats(data=daten_short, 
                        x=Asthma, 
                        y=Genetische_PräAs,
                        label='both',
                        digits=3, 
                        proportion.test=FALSE, # 1st Edit
                        # became clear after OP's comment below this answer
                        bf.message=FALSE # 2nd Edit
                        ) |> # 4th Edit (fully piped)
  (\(.) {.@labels$subtitle = as.call(as.list(.@labels$subtitle)[c(3, 7)]); .})()

2nd Edit.

Note that iff bf.message=FALSE (default is TRUE) is set we can pass custome input to caption. As it defaults to NULL we do not need to overwrite caption in downstream.

caption
The text for the plot caption. This argument is relevant only if bf.message = FALSE.

proportion.test defaults to results.subtitle. If you run this test over and over we could consider modifying that part of the function (with welcoming side effects).

3th Edit.

DIY, for the fun (and education)

# statistics 
tab = with(daten_short, table(Asthma, Genetische_PräAs))
rb = prop.table(tab, margin=2)
C = adjustcolor(c('#CE4A08', '#1D8F64'), alpha.f=.8)
m = bquote(italic(p)==.(round(stats::chisq.test(tab, correct=FALSE)$p.value, 3)) 
            ~ ',' ~ italic(n)[obs]==.(nrow(daten_short)))
# barplot
b = barplot(rb, yaxt='n', col=C, xlab=bquote(bold('Genetische_PräAs')), main=m)
axis(2, s <- seq(0, 1, .1), sprintf('%g%%', s*100), las=2)
legend('bottom', legend = rownames(rb), fill=C, bty='n', title='Asthma')
# labels 
ym = apply(rb, 2, cumsum) - rb/2L
text(rep(b, each=nrow(rb)), ym, labels=sprintf('%s\n%.1f%%', tab, rb*100))

White label boxes can be added with rect where one uses b and nr to calculate positions. For legend positions we recommend to search for it on this platform.

edited Nov 26 at 20:23

answered Nov 25 at 15:29

Friede

11.8k2 gold badges14 silver badges32 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

André · Accepted Answer · 2025-11-25 19:20:12Z

-1

See the documentation for how to adapt the plot. You can try the following which alters the subtitle and removes caption a posteriori:

daten_short<-structure(list(Asthma = c(0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 
                                       0, 0, 0, 1, 0, 0, 1, 1), Genetische_PraAs = c("Nein", "Nein", 
                                                                                     "Nein", "Nein", "Ja", "Nein", "Ja", "Nein", "Ja", "Nein", "Nein", 
                                                                                     "Nein", "Nein", "Nein", "Nein", "Nein", "Ja", "Nein", "Nein", 
                                                                                     "Ja")), row.names = c(NA, -20L), class = c("tbl_df", "tbl", "data.frame"
                                                                                     ))

library(ggstatsplot)

# Create plot with p-value 
plot <- ggbarstats(
  data            = daten_short,
  x               = Asthma,
  y               = Genetische_PraAs,
  label           = "both",             # counts and percentage
  proportion.test = FALSE,              # remove p-values from the top
  digits           = 3,
)

# Add custom subtitle and empty caption
plot + labs(
  subtitle = bquote(.(plot$labels$subtitle[[3]]) * ", " * .(plot$labels$subtitle[[7]])),
  caption  = NULL
)

This produces the following:

edited Nov 25 at 19:20

answered Nov 25 at 12:56

André

1,98614 silver badges28 bronze badges

10 Comments

Limey Nov 25 at 13:20

This answer could be improved by adding the result of running the suggested solution to demonstrate that it meets OP's needs.

Sourish Rakshit Nov 25 at 13:29

Sorry but it is getting rid of all results and percentages. I want to keep the percentages and counts and the p-value as well.

André Nov 25 at 14:03

I would love to, need the data. OP should post a minimal reproducible example.

Sourish Rakshit Nov 25 at 14:27

Just to clarify. I dont want to get rid of all stats. I want to keep the p value and the count of observed values (nobs) on top. I just want to get rid of all the others sich as cramers V

Sourish Rakshit Nov 25 at 14:25

@André I have edited the post. You will be able to execute it in R with a small sample data. Thanks!

Friede Nov 25 at 16:24

The essential part ot this answer is as.list() |> ( \(.) c(as.name('list'), .[[3]], .[[7]]) )() |> as.call() which is simply copied over from mine. I therefore flagged and downvoted this answer.

André Nov 25 at 18:54

I disagree, removing the individual p= as well as the caption also was, see the comment threads. We both edited to achieve the desired output and I referenced copying from your answer. A pity that you take that personal., no offense intended...

|

Collectives™ on Stack Overflow

How do I erase these labels in ggbarstats plot?

2 Answers 2

2 Comments

10 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

10 Comments

Your Answer

Sign up or log in

Post as a guest

Related