0

I need to label each bar from d_ranked_long (height = rank) with its rank's absolute value in d_long$value, so I added geom_text(data = d_long, aes(x=parameter, label=value)) to the ggplot at the end but got this error.

Error in `f()`:
! Aesthetics must be valid data columns. Problematic aesthetic(s): y = rank. 
Did you mistype the name of a data column or forget to add after_stat()?
Backtrace:
 1. base `<fn>`(x)
 2. ggplot2:::print.ggplot(x)
 4. ggplot2:::ggplot_build.ggplot(x)
 5. ggplot2 by_layer(function(l, d) l$compute_aesthetics(d, plot))
 6. ggplot2 f(l = layers[[i]], d = data[[i]])
 7. l$compute_aesthetics(d, plot)
 8. ggplot2 f(..., self = self)

enter image description here

With the geom_text() command, as solved here How to put labels over geom_bar in R with ggplot2, I'm telling R: with each (x,y) in d_long_ranked, grab the value in the corresponding position in d_long and put it on top of the bar. For example, the first bar in the first panel (vs = 0, am = 0) should have 3.44 on top of it, the second bar in the first panel should have 8.00 on top of it, and so on.

sample data

d_wide <- head(mtcars) %>%
  rownames_to_column(var = "make_model") %>%   select(!c(make_model,gear, carb))

absolute values in long format

d_long <-  d_wide  %>%
  pivot_longer(!c(vs, am), names_to = "parameter", values_to = "value")

Save the ranks in a separate dataframe

d_ranked <-  data.frame(d_wide [,c(8:9)],
           t(apply(d_wide[,-c( 8:9)], 1, 
                   rank, ties.method='min')))

the ranks in long format

d_ranked_long <- d_ranked  %>%
  pivot_longer(!c(vs, am), names_to = "parameter", values_to = "rank")

The x-axis ticks need to be in the following order:

d_ranked_long$parameter <- factor(d_ranked_long$parameter, levels = c("wt", "cyl", "drat", "hp",
                                                        "disp", "qsec","mpg"))

d_long$parameter <- factor(d_long$parameter, levels = c("wt", "cyl", "drat", "hp",
                                                        "disp", "qsec","mpg"))

put everything together

d_ranked_long %>%
ggplot(aes(x=parameter, y=rank)) +
  geom_bar(stat = "identity") +
  geom_text(data = d_long, aes(x=parameter, label=value)) +
  facet_grid(am ~ vs) +
  theme(axis.text.x = element_text(angle = 90))
hnguyen
  • 772
  • 6
  • 17
  • 3
    The problem is that `geom_text` is inheriting aes `y = rank` but `rank` is not a `d_long` column. You need to revise your data wrangling code. – Rui Barradas Jun 06 '22 at 07:37
  • 2
    There is no reason why you have two data sets, why not `d <- cbind(d_ranked_long, value= d_long$value)` and plot `d`? I'm only seeing a problem, the lower left plot has 2 labels per bar. – Rui Barradas Jun 06 '22 at 07:53

0 Answers0