Help with R markdown to word: flextables with results = 'asis'

Publishing Your Results
,
1

Hi, I have an issue with this markdown code.

I have tried different options but the word document does not show the tables I expect.

Here is the reprex of the Markdown document .

The codes in chunks iris.3 to iris.5 works when I run the chuck but don't work when I knit it to Word.
They are different options I have seen:

  • in chunk iris.3: walk()
  • in chunk iris.4: flextable_to_rmd()
  • in chunk iris.5: knit_print()

I will add the reprex of the individual chunks afterwards.
Many thanks in advance for any help.

---
title: "example"
author: "Beatriz"
date: "`r Sys.Date()`"
output: word_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

library(knitr)
library(rstatix)
library(tidyverse)
library(tidyxl)
library(lubridate)
library(skimr)
library(flextable)
```

1. This example works

```{r iris.1}

iris %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
       

```

2. This example works

```{r iris.2}

tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

tabulate_by_spp("setosa") # works        

```

3. This example works when I run the chuck in R but the Word document does not show the tables.

It uses **walk()** which i have used successfully to print plots (ggplots) using very similar functions.

```{r iris.3,  results='asis'}

tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

plots <-
  levels(iris$Species) %>% 
  map(tabulate_by_spp) 

plots %>% 
  walk(print)

```

4. This example works when I run the chuck in R but the Word document does not show the tables.

It uses **flextable_to_rmd()** instead of **walk()** 

```{r iris.4,  results='asis'}

tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

plots <-
  levels(iris$Species) %>% 
  map(tabulate_by_spp) 

flextable_to_rmd(plots)

```

5. This example works when I run the chuck in R but the Word document does not show the tables.

It uses **knit_print()** instead of **walk()**  

```{r iris.5,  results='asis'}

tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

plots <-
  levels(iris$Species) %>% 
  map(tabulate_by_spp) 

knit_print(plots)

```

#> Error in parse(text = input): attempt to use zero-length variable name

Created on 2025-12-02 with reprex v2.1.1

Individual chunks:

iris.1. This example works:

library(knitr)
library(rstatix)
#> 
#> Attaching package: 'rstatix'
#> The following object is masked from 'package:stats':
#> 
#>     filter
library(tidyverse)
library(tidyxl)
library(lubridate)
library(skimr)
library(flextable)
#> Warning: package 'flextable' was built under R version 4.4.3
#> 
#> Attaching package: 'flextable'
#> The following object is masked from 'package:purrr':
#> 
#>     compose
iris %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R)

Created on 2025-12-02 with reprex v2.1.1

iris.2. This example works

library(knitr)
library(rstatix)
#> 
#> Attaching package: 'rstatix'
#> The following object is masked from 'package:stats':
#> 
#>     filter
library(tidyverse)
library(tidyxl)
library(lubridate)
library(skimr)
library(flextable)
#> Warning: package 'flextable' was built under R version 4.4.3
#> 
#> Attaching package: 'flextable'
#> The following object is masked from 'package:purrr':
#> 
#>     compose
tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

tabulate_by_spp("setosa") # works

Created on 2025-12-02 with reprex v2.1.1

iris.3. This example works when I run the chuck in R but the Word document does not show the tables.

It uses walk() which i have used successfully to print plots (ggplots) using very similar functions.

library(knitr)
library(rstatix)
#> 
#> Attaching package: 'rstatix'
#> The following object is masked from 'package:stats':
#> 
#>     filter
library(tidyverse)
library(tidyxl)
library(lubridate)
library(skimr)
library(flextable)
#> Warning: package 'flextable' was built under R version 4.4.3
#> 
#> Attaching package: 'flextable'
#> The following object is masked from 'package:purrr':
#> 
#>     compose
tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

plots <-
  levels(iris$Species) %>% 
  map(tabulate_by_spp) 

plots %>% 
  walk(print)
#> a flextable object.
#> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> header has 1 row(s) 
#> body has 1 row(s) 
#> original dataset sample: 
#> 'data.frame':    1 obs. of  8 variables:
#>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 1
#>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#>  $ n       : num 50
#>  $ min     : num 4.3
#>  $ max     : num 5.8
#>  $ q1      : num 4.8
#>  $ median  : num 5
#>  $ q3      : num 5.2
#> a flextable object.
#> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> header has 1 row(s) 
#> body has 1 row(s) 
#> original dataset sample: 
#> 'data.frame':    1 obs. of  8 variables:
#>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 2
#>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#>  $ n       : num 50
#>  $ min     : num 4.9
#>  $ max     : num 7
#>  $ q1      : num 5.6
#>  $ median  : num 5.9
#>  $ q3      : num 6.3
#> a flextable object.
#> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> header has 1 row(s) 
#> body has 1 row(s) 
#> original dataset sample: 
#> 'data.frame':    1 obs. of  8 variables:
#>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 3
#>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#>  $ n       : num 50
#>  $ min     : num 4.9
#>  $ max     : num 7.9
#>  $ q1      : num 6.22
#>  $ median  : num 6.5
#>  $ q3      : num 6.9

Created on 2025-12-02 with reprex v2.1.1

iris.4. This example works when I run the chuck in R but the Word document does not show the tables.

It uses flextable_to_rmd() instead of walk()

library(knitr)
library(rstatix)
#> 
#> Attaching package: 'rstatix'
#> The following object is masked from 'package:stats':
#> 
#>     filter
library(tidyverse)
library(tidyxl)
library(lubridate)
library(skimr)
library(flextable)
#> Warning: package 'flextable' was built under R version 4.4.3
#> 
#> Attaching package: 'flextable'
#> The following object is masked from 'package:purrr':
#> 
#>     compose


tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

plots <-
  levels(iris$Species) %>% 
  map(tabulate_by_spp) 

flextable_to_rmd(plots)

#> 
#> 
#> ```
#> #> [[1]]
#> #> a flextable object.
#> #> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> #> header has 1 row(s) 
#> #> body has 1 row(s) 
#> #> original dataset sample: 
#> #> 'data.frame': 1 obs. of  8 variables:
#> #>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 1
#> #>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#> #>  $ n       : num 50
#> #>  $ min     : num 4.3
#> #>  $ max     : num 5.8
#> #>  $ q1      : num 4.8
#> #>  $ median  : num 5
#> #>  $ q3      : num 5.2
#> #> 
#> #> [[2]]
#> #> a flextable object.
#> #> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> #> header has 1 row(s) 
#> #> body has 1 row(s) 
#> #> original dataset sample: 
#> #> 'data.frame': 1 obs. of  8 variables:
#> #>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 2
#> #>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#> #>  $ n       : num 50
#> #>  $ min     : num 4.9
#> #>  $ max     : num 7
#> #>  $ q1      : num 5.6
#> #>  $ median  : num 5.9
#> #>  $ q3      : num 6.3
#> #> 
#> #> [[3]]
#> #> a flextable object.
#> #> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> #> header has 1 row(s) 
#> #> body has 1 row(s) 
#> #> original dataset sample: 
#> #> 'data.frame': 1 obs. of  8 variables:
#> #>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 3
#> #>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#> #>  $ n       : num 50
#> #>  $ min     : num 4.9
#> #>  $ max     : num 7.9
#> #>  $ q1      : num 6.22
#> #>  $ median  : num 6.5
#> #>  $ q3      : num 6.9
#> #> 
#> #> $caption
#> #> $caption$word_stylename
#> #> [1] "Table Caption"
#> ```

Created on 2025-12-02 with reprex v2.1.1

iris.5. This example works when I run the chuck in R but the Word document does not show the tables.

It uses knit_print() instead of walk()

library(knitr)
library(rstatix)
#> 
#> Attaching package: 'rstatix'
#> The following object is masked from 'package:stats':
#> 
#>     filter
library(tidyverse)
library(tidyxl)
library(lubridate)
library(skimr)
library(flextable)
#> Warning: package 'flextable' was built under R version 4.4.3
#> 
#> Attaching package: 'flextable'
#> The following object is masked from 'package:purrr':
#> 
#>     compose

tabulate_by_spp <- function(Species_name) {

  table_by_spp <-
    iris %>%
    filter(Species == Species_name)  %>%
  group_by(Species) %>%
  get_summary_stats(Sepal.Length, type = "five_number")%>% # this is necessary to print as image
  flextable::flextable() %>%    # convert to image
  flextable::autofit()   %>%    
  flextable::fit_to_width(max_width = 7) # Fits the table to max width for Word(R) 
 

 return(table_by_spp)

}

plots <-
  levels(iris$Species) %>% 
  map(tabulate_by_spp) 

knit_print(plots)
#> [[1]]
#> a flextable object.
#> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> header has 1 row(s) 
#> body has 1 row(s) 
#> original dataset sample: 
#> 'data.frame':    1 obs. of  8 variables:
#>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 1
#>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#>  $ n       : num 50
#>  $ min     : num 4.3
#>  $ max     : num 5.8
#>  $ q1      : num 4.8
#>  $ median  : num 5
#>  $ q3      : num 5.2
#> 
#> [[2]]
#> a flextable object.
#> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> header has 1 row(s) 
#> body has 1 row(s) 
#> original dataset sample: 
#> 'data.frame':    1 obs. of  8 variables:
#>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 2
#>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#>  $ n       : num 50
#>  $ min     : num 4.9
#>  $ max     : num 7
#>  $ q1      : num 5.6
#>  $ median  : num 5.9
#>  $ q3      : num 6.3
#> 
#> [[3]]
#> a flextable object.
#> col_keys: `Species`, `variable`, `n`, `min`, `max`, `q1`, `median`, `q3` 
#> header has 1 row(s) 
#> body has 1 row(s) 
#> original dataset sample: 
#> 'data.frame':    1 obs. of  8 variables:
#>  $ Species : Factor w/ 3 levels "setosa","versicolor",..: 3
#>  $ variable: Factor w/ 1 level "Sepal.Length": 1
#>  $ n       : num 50
#>  $ min     : num 4.9
#>  $ max     : num 7.9
#>  $ q1      : num 6.22
#>  $ median  : num 6.5
#>  $ q3      : num 6.9

Created on 2025-12-02 with [reprex v2.1.1]

2

Hi all,

I have seen that the post had some reads but no answers.

Is there any way I can improve my question?

Reading more over the topic I have seen posts in which following the use of loops the flextables withing the chunk are printed in Word documents are printed with no issues when in the chuck options includes results = 'asis'.

I have this in my code but doesn't work with it.

However if I replace flextable() for kable() it prints out the tables in Word. I am properly puzzled ... and it is likely a silly coding issue... likely because plots is not a data.frame but a list?

Happy to hear suggestions :slight_smile:

3

Hi again, I found a solution using a child document as described here.

under section 4.3.2 Looping in R Markdown documents

If anyone if interested I can share the code when I refine it. If not I will let this one be closed if there are no replies to this.

B