Boxplot with ggplot2 in R - returns by month

Question

I have computed monthly returns from a price series. I then build a dataframe as follows:

    y.ret_1981    y.ret_1982    y.ret_1983   y.ret_1984   y.ret_1985
1   0.0001015229  0.0030780203 -0.0052233836  0.017128325 -0.002427308
2   0.0005678989  0.0009249838 -0.0023294622 -0.030531971  0.001831160
3  -0.0019040392 -0.0021614791  0.0022451252 -0.003345983  0.005773503
4  -0.0006015118  0.0010695681  0.0052680258  0.008592513  0.009867972
5   0.0052736054 -0.0003181347 -0.0008505673 -0.000623061 -0.012225140
6   0.0014266119 -0.0101045071 -0.0003073150 -0.016084505 -0.005883687
7  -0.0069002733 -0.0078170620  0.0070058676 -0.007870294 -0.010265335
8  -0.0041963258  0.0039905142  0.0134996961 -0.002149331 -0.007860940
9   0.0020778541 -0.0038834826  0.0052289589  0.007271409 -0.005320848
10  0.0030956487 -0.0005027686 -0.0021452210  0.002502301 -0.001890657
11 -0.0032375542  0.0063916686  0.0009331531  0.004679741  0.004338580
12  0.0014882164  0.0039578527  0.0136663415  0.000000000  0.003807668

... where columns are the monthly returns for the years 1981 to 1985 and the rows 1 to 12 are the months of the year.

I would like to plot a a boxplot similar to the one below:

So what can I go? And I would like my graph to read the months of the years instead of integer 1 to 12.

Thank you.

Answer 1

First, add new column month to your original data frame containing month.name (built-in constant in R) and use it as factor. It is import to set also levels= inside the factor() to ensure that months are arranged in chronological order not the alphabetical.

Then melt this data frame from wide format to long format. In ggplot() use month as x values and value as y.

df$month<-factor(month.name,levels=month.name)
library(reshape2)
df.long<-melt(df,id.vars="month")
ggplot(df.long,aes(month,value))+geom_boxplot()