Tell me more ×
Stack Overflow is a question and answer site for professional and enthusiast programmers. It's 100% free, no registration required.
up vote -1 down vote favorite

I have a simple data frame called msq:

    sex wing    index
1   h   54      67.4
2   m   60.5    67.9
3   m   60      64.5
4   m   59      66.6
5   m   63.5    63.3
6   m   63      66.7
7   m   61.5    71.8
8   m   62      67.9
9   m   63      67.8
10  m   62.5    72.7
11  m   61.5    70.3
12  h   54.5    70.7
13  m   60      61.1
14  m   63.5    50.9
15  m   63      72.1

My intention is to make a boxplot with ggplot for which I use this code that works fine:

gplot(msq, aes("index",index))+ geom_boxplot (aes(group="sex"))

and then to plot an outlier that should stand alone up in the graph (a value 73.9). The problem is that if I include it in the data set, the boxplot "absorbs" it making the error line longer... I have been looking in Hmisc and to stat_summary but I can't get any clear idea.

thank you.

share|improve this question
1  
What's your question here? What are you trying the box-plot against? Only sex vs index? Also you should remove the "" in sex if you want to group them by it. – Arun Feb 3 at 8:23
I just need to add a point in y=73.8.Yes, thanks for the "" comment... – Miguel Tirado Feb 3 at 9:14

1 Answer

up vote 1 down vote accepted

You could use geom_point to add points to a plot generated with ggplot2.

library(ggplot2)
ggplot(msq, aes(sex, index)) +   # Note. I modified the aes call
  geom_boxplot() +
  geom_point(aes(y = 73.9)) # add points

enter image description here

share|improve this answer
1  
Is it right to do that? – Arun Feb 3 at 9:38
@Arun I suppose it's better to use geom_point. I modified the answer. – Sven Hohenstein Feb 3 at 9:49
1  
I did not mean the command. annotate and geom_point does the job for sure. But do you think it is appropriate to decide the outliers? Although I understand that's what the OP is asking for. Just wanted to know what others think in this matter. – Arun Feb 3 at 9:52
1  
@Arun I fully agree. This is not the right way to deal with outliers. But since the OP wants to avoid longer boxplot lines (what I think is incorrect), there seems to be not other way. – Sven Hohenstein Feb 3 at 10:14
Thank you! it is exactly what I needed. I know this is not the normal way to plot an outlier, but the thing is that 73.9 is a single observation I think belongs to another population. – Miguel Tirado Feb 3 at 10:48

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.