How to export information from R objects as png?

Question

I want to summarize information about my data and export it as a png file. I've experimented with packages such as pander and flextable, but currently can't achieve my goal. Maybe there's a way to get this done using ggplot2, but I'm unaware of such.

Example

Let's say that we have the mtcars data, and we want to extract some information about it:

• Number of rows in the data
• Average of mpg
• The factor levels available in cyl
• Regression summary for predicting mpg ~ cyl

To this end, I'll compute each of the above and assign them to objects. Finally, I'll bundle all the info in a list object.

library(dplyr)
library(tibble)
library(broom)

number_of_rows <- nrow(mtcars)
mpg_mean       <- mean(mtcars$mpg)
cyl_levels     <- mtcars %>% select(cyl) %>% unique() %>% remove_rownames()
model_summary  <- lm(mpg ~ cyl, mtcars) %>% broom::tidy()

my_data_summary <- lst(number_of_rows,
                       mpg_mean,
                       cyl_levels,
                       model_summary)


> my_data_summary
## $number_of_rows
## [1] 32

## $mpg_mean
## [1] 20.09062

## $cyl_levels
##   cyl
## 1   6
## 2   4
## 3   8

## $model_summary
## # A tibble: 2 x 5
##   term        estimate std.error statistic  p.value
##   <chr>          <dbl>     <dbl>     <dbl>    <dbl>
## 1 (Intercept)    37.9      2.07      18.3  8.37e-18
## 2 cyl            -2.88     0.322     -8.92 6.11e-10

---

My question: how can I export my_data_summary as a png?

---

Some unsuccessful attempts.

To note, I'm not particularly motivated to use any specific package.

1) I've tried pander:

library(pander)

pander(my_data_summary)



  * **number_of_rows**: _32_
  * **mpg_mean**: _20.09_
  * **cyl_levels**:

    -----
     cyl
    -----
      6

      4

      8
    -----

  * **model_summary**:

    ------------------------------------------------------------
        term       estimate   std.error   statistic    p.value
    ------------- ---------- ----------- ----------- -----------
     (Intercept)    37.88       2.074       18.27     8.369e-18

         cyl        -2.876     0.3224       -8.92     6.113e-10
    ------------------------------------------------------------


<!-- end of list -->

This indeed gets me pretty far, but insufficient. How can I get from such markdown textual output to a rendered png?

---

2) I've also tried flextable.

library(flextable)

flextable(my_data_summary)

Error in flextable(.) : is.data.frame(data) is not TRUE

OK, so flextable() accepts only data.frame class.

I therefore could have done something like:

flextable(my_data_summary$model_summary)

Which then gives this nice output that can be saved to png with flextable::save_as_image():

However, the problem with flextable(my_data_summary$model_summary) is that I want to get the entire contents of my_data_summary rendered and exported to that single png, similar to how pander() accepts the list object in its entirety.

Also important to note that I execute this code via Rscript, so I need a solution that is not interactive/GUI-based.

---

Desired Output

A png file that represents my_data_summary object and looks something like:

Any ideas on this?

---

EDIT

---

Based on @Waldi's answer below, I'd like to emphasize that I'm looking for a solution that could be run in a single code run, without creating intermediate temporary files. In other words, I'm trying to come up with a code that has a single output: a png file of my_data_summary contents.

Answer 1

You can use htmlTable package which has the function to concatenate multiple html tables

library(dplyr)
library(tibble)
library(broom)
number_of_rows <- nrow(mtcars)
mpg_mean       <- mean(mtcars$mpg)
cyl_levels     <- mtcars %>% select(cyl) %>% unique() %>% remove_rownames()
model_summary  <- lm(mpg ~ cyl, mtcars) %>% broom::tidy()

my_data_summary <- lst(number_of_rows,
                       mpg_mean,
                       cyl_levels,
                       model_summary)


#Using htmlTable
library(htmlTable)

#converting each element of the list into invidual html table and concatenating 
concatHtmlTables(list(htmlTable(my_data_summary[[1]]), 
                  htmlTable(txtRound(my_data_summary[[2]], 4)),
                  htmlTable(my_data_summary[[3]]),
                  htmlTable(txtRound(my_data_summary[[4]], 4))), 
             headers = names(my_data_summary))

      

Answer 2

You could use the webshot package.
As webshot relies on phantom.js the first step is to run:

webshot::install_phantomjs()

Then create test.Rmd file :

---
title: test
output: html_document
---

`r knitr::opts_chunk$set(echo = FALSE, warning = FALSE, message = FALSE, cache = F)`

```{r,results='asis'}
library(dplyr)
library(tibble)
library(broom)

number_of_rows <- nrow(mtcars)
mpg_mean       <- mean(mtcars$mpg)
cyl_levels     <- mtcars %>% select(cyl) %>% unique() %>% remove_rownames()
model_summary  <- lm(mpg ~ cyl, mtcars) %>% broom::tidy()

my_data_summary <- lst(number_of_rows,
                       mpg_mean,
                       cyl_levels,
                       model_summary)

library(pander)

pander(my_data_summary)
```

You are now ready to output png:

webshot::rmdshot('test.Rmd','test.png')