When months have unique labels

The other day I was working on a project which labelled months unusually: Month 1 was October, 2 was November etc. Because the analysis involved rasters it was important to keep the numbering system. However, when I changed the raster into dataframes it was better for graphing in ggplot to have the numbers replaced by the names of the months. While the tidyverse is great most of the time for a situation like this it is often easiest to use base R.

Making the mock data

Let’s make some mock data which emulates this situation.

# a reference which corresponds to the order
# of the months (but not their name)
reference <- 1:9

# some unimportant data 
category <- letters[1:9]

# make a data frame
df <- data.frame(reference , category)

Now df looks like:

  reference category
1         1        a
2         2        b
3         3        c
4         4        d
5         5        e
6         6        f
7         7        g
8         8        h
9         9        i

Next I use the month.names[] constant to produce the sequence of numbers which corresponds to the months:

ref_names <- c(month.name[10:12], month.name[1:9])
ref_names

Which results in:

 [1] "October"   "November"  "December"  "January"  
 [5] "February"  "March"     "April"     "May"      
 [9] "June"      "July"      "August"    "September"

Now the tricky bit, there are more months in the sequence than there are actual monitoring months. To ensure they match we use the following code, cleverly using the max() function to ensure we only include matching months:

my.months <-ref_names[1:max(df$reference)]
my.months 

Which results in:

[1] "October"  "November" "December" "January" 
[5] "February" "March"    "April"    "May"     
[9] "June" 

Next we change reference to a factor. We then use the levels() function to ensure the factor levels are those given by my.month.

df$reference <- as.factor(df$reference)
levels(df$reference) <- factor(my.months)

Now let’s check that it worked by calling df:

df

    reference category
1   October        a
2  November        b
3  December        c
4   January        d
5  February        e
6     March        f
7     April        g
8       May        h
9      June        i

Hooray…it works!

BACK to INDEX