Passing column name as parameter to R function

Question

I have a piece of code which produces a graph in R. It works, but there's a field name star_sign which is hard coded in the function. I'd like to pass this as an extra function parameter variable_name instead and remove the hard coding.

my_graph <- function (df_name, variable_name)
{
  scale_fac <- 4 * max(df_name$exposure) / max(df_name$max) # rule of thumb - it'll do for now

  row_count = nrow(df_name)

  #problem line here - star_sign appears twice

  df_name <- df_name %>% mutate(star_sign = as.numeric(factor(star_sign, levels = my_levels)))

  df_name %>%
    ggplot(aes_string(x = variable_name)) +
    geom_line(aes(y = mean), color = "red") +
    geom_line(aes(y = min), color = "blue") +
    geom_line(aes(y = max), color = "green") +
    geom_col(aes(y = exposure / scale_fac), width = 0.5, fill = "blue") +
    scale_y_continuous("Linear Predictor", sec.axis = sec_axis(~ .*scale_fac , name = "Exposure")) +
    scale_x_continuous(variable_name, breaks = c(1:row_count), labels = df_name$my_levels) +
    theme_bw()
}

df <- data.frame(star_sign = c("Aries", "Taurus", "Gemini", "Cancer", "Leo", "Virgo", "Libra", "Scorpio", "Sagittarius", "Capricorn", "Aquarius", "Pisces"),
                 exposure = c(50, 70, 60, 40, 45, 78, 42, 22, 28, 49, 50, 31),
                 mean = c(1.1, 1.2, 1.4, 1.3, 1.8, 1.6, 1.4, 1.3, 1.2, 1.1, 1.5, 1.3))

df$min <- 0.95 * df$mean
df$max <- 1.05 * df$mean

df$my_levels = c("Aries", "Taurus", "Gemini", "Cancer", "Leo", "Virgo", "Libra", "Scorpio", "Sagittarius", "Capricorn", "Aquarius", "Pisces")

my_graph(df, "star_sign")

The hard coding is in this line

  df_name <- df_name %>% mutate(star_sign = as.numeric(factor(star_sign, levels = my_levels)))

I've read various SO articles about passing column names to R functions but I'm still a relative beginner with R and I'm really struggling to follow them. Can anyone advise the syntax I need please?

Thank you.

Answer 1

We can convert the string input to sym bol and then do the evaluation ( !! )

my_graph <- function (df_name, variable_name)
{
  scale_fac <- 4 * max(df_name$exposure) / max(df_name$max) # rule of thumb - it'll do for now

  row_count = nrow(df_name)

  #problem line here - star_sign appears twice

  df_name <- df_name %>% mutate(!! variable_name := as.numeric(factor(!! rlang::sym(variable_name), levels = my_levels)))

  df_name %>%
    ggplot(aes(x = !! rlang::sym(variable_name))) +
    geom_line(aes(y = mean), color = "red") +
    geom_line(aes(y = min), color = "blue") +
    geom_line(aes(y = max), color = "green") +
    geom_col(aes(y = exposure / scale_fac), width = 0.5, fill = "blue") +
    scale_y_continuous("Linear Predictor", sec.axis = sec_axis(~ .*scale_fac , name = "Exposure")) +
    scale_x_continuous(variable_name, breaks = c(1:row_count), 
             labels = df_name$my_levels) +
    theme_bw()
}

-checking

my_graph(df, "star_sign")