Insert image with kableExtra function::spec_image( )

Question

library(tidyverse)
library(kableExtra)

Suppose the following dataset:

games <- tibble(logo = "",
                nome = c("Joao", "Pedro", "Geovanna"),
                medalha = c("Gold", "Silver","Bronze"))

I want to add pictures (.png) of the corresponding medals. I created a PROJECT (File > New Project...) to work with this data so as not to worry about setting the complete path to the files.

So I have a folder called "img" where the following files are:

• 1024px-Bronze_medal.png
• 1024px-Gold_medal.png
• 1024px-Silver_medal.png

To insert the images of the medals, I did the following:

games %>% 
  kable() %>% 
  kable_paper(full_width = T) %>% 
  column_spec(1, image = spec_image(c("img/1024px-Gold_medal.png",
                                      "img/1024px-Silver_medal.png",
                                      "img/1024px-Bronze_medal.png"), 200,200))

When I did the Knit the result was as expected:

The drawback was that I needed to inform each of the paths in the function spec_image().

I’m trying to optimize this process. For this, I imagined the following: use the function sprintf() to create a new column in the dataset (which I called "link". I worried about inserting the quotation marks) with the paths and pass this newly created column as an argument to the function spec_image()

games %>% 
  mutate(link = sprintf("'img/1024px-%s_medal.png'", medalha)) %>% 
  kable() %>% 
  kable_paper(full_width = T) %>% 
  column_spec(1, image = spec_image("link", 200,200))

However, when doing Knit the following error message appears:

File link not found in resource path
Erro: pandoc document conversion failed with error 99

I thought when referencing the "link" column in the function spec_image() would be the same thing spec_image(c("img/1024px-Gold_medal.png", "img/1024px-Silver_medal.png", "img/1024px-Bronze_medal.png"))

Answer 1

I solved the problem, but perhaps not the answer sought.

(reread the images of the medals. as they were not provided in the original post, I searched others on the internet)

The trick I used was to create the object link outside the pipe, so as not to confuse the dplyr and the kable.

library(knitr)
library(kableExtra)
library(tidyverse)

games <- tibble(logo = "",
                nome = c("Joao", "Pedro", "Geovanna"),
                medalha = c("Gold", "Silver","Bronze"))

medalha <- c("Gold", "Silver", "Bronze")

link <- sprintf("img/1024px-%s_medal.png", medalha)

kable(games) %>% 
  kable_paper(full_width = T) %>% 
  column_spec(1, image = spec_image(link, 200, 200))

I understand that, although it is working, this may not be the solution sought by the AP. However, in a post on the Rstudio forum entitled Conditional formatting with column_spec Within a dplyr chain, Hao Zhu, the author of the package kableExtra, solves the problem that is put there in a similar way. He even comments the following:

The Solution is quite simple: save the final data before you create your table. Well, I understand the Joy of Piping from top to bottom. However, in cases like this one, it’s just easier and Cleaner to break them into two Pieces.

In a free translation,

The solution is quite simple: save the final dataset before creating your table. Well, I understand the satisfaction of creating a pipe from beginning to end of your code. However, in cases like this, it’s just easier and cleaner to break them into two parts.