Question

I have a dataframe(df) with 3 columns:

test1   test2    res
500    23        30.75
200    23        31.36
110    26        30.38
105    23        31.57
55     21        33.56
35     21        31.39
27     18        36.89
23     23        35.44 

I want to have one graph contians boxplots for both results. I have melted the data based on “res” and called the new one (df_melt) and I got the following:

res      variable  value
30.75     test1   500    
31.36     test1   200
30.38     test1   110
31.57     test1   105
33.56     test1    55
31.39     test1    35
36.89     test1    27
35.44     test1    23
30.75     test2    23
31.36     test2    23
30.38     test2    26
31.57     test2    23
33.56     test2    21
31.39     test2    21
36.89     test2    18
35.44     test2    23 

Now, I want to boxplot the data so, x-axis=res, and y-axis= value based on variable column. Therefore, I would expect two boxplots for each “res” value as I have two type of data in variable ( i.e.test1 and test2).

I have tried this :

boxplot(df_melt$value~df_melt$res)

I got only 8 boxplot and it is not clear for which test??? How I can add the second test to the graph so I can compare them.

The whole reason why I want to include them together so I can compare them ….

Answer 1

I suppose you are looking for a barplot, since you can’t generate a boxplot based on a single value.

library(ggplot2)

ggplot(df_melt, aes(x = factor(res), y = value, fill = variable)) +
  geom_bar(stat="identity", position="dodge")

---

Update: A boxplot solution based on the comment of the OP:

An analogous graph with multiple boxplots in different colours can be generated with:

ggplot(df_melt, aes(x = factor(res), y = value, fill = variable)) +
  geom_boxplot() +
  scale_fill_manual(values = c("white", "darkgray"))

Note that this graph only makes sense if you have multiple values in values for each combination of res and variable.

The colours of the boxplots will be white and dark grey. You can modify the colour with the scale_fill_manual function.