tableboom

This package received Honorable Mention in Posit Table Contest 2022.

The goal of {tableboom} is to let user choose the R script to inspect intermediate results of all calls (using {boomer}) in the form of HTML table. Three main features are:

• give the user possibility to inspect intermediate steps of all calls in the chosen R script, i.e. no need to boomer::boom() each call separately;
• prepare output as an HTML document, so it can be read outside of programming process;
• prepare HTML document as a table, but trying to make it more like a word story than table;

With this software, vignette Treating Code As Data - Notes was prepared as some kind of tutorial about metaprogramming (in the part it is used in {tableboom}), which is available also online: Treating Code As Data - Notes.

Installation

You can install the development version of {tableboom} from GitHub with:

# install.packages("devtools")
devtools::install_github("gsmolinski/tableboom")

# or with vignette, but installation time will be much longer
devtools::install_github("gsmolinski/tableboom", build_vignettes = TRUE)

Example

Having the following example script which we would like to inspect:

library(dplyr)

path <- file.path(system.file(package = "tableboom","table_contest_2022", "inner_script"),
                  "eurostat_data.R")
# Unfortunately, comments in separate lines are not included
source(path) # data - children (< 18) from Ukraine which found temp protection

names(child_ukr_prot) <- c("country", "march_2022", "april_2022",
                           "may_2022","june_2022", "july_2022", "august_2022")

child_ukr_prot[5, "country"] <- stringi::stri_replace_all_regex(child_ukr_prot[5, "country"],
                                                                "\\s+.+", "")

child_ukr_prot <- child_ukr_prot |>
  mutate(across(matches("\\d$"), \(e) if_else(e == ":", NA_character_, e)),
         across(matches("\\d$"), as.integer))

maxs <- vector("integer", ncol(child_ukr_prot) - 1)

for (i in 2:(ncol(child_ukr_prot))) {
  maxs[[i - 1]] <- max(child_ukr_prot[[i]], na.rm = TRUE)
}

maxs

total_months <- colSums(child_ukr_prot[, 2:ncol(child_ukr_prot)], na.rm = TRUE)

format(sum(total_months), big.mark = " ")

path()

the usage of {tableboom} is nothing more than calling inspect_r() with the path to the script (or NULL argument - and then saved and opened script in the RStudio editor will be used):

path <- file.path(system.file(package = "tableboom", "table_contest_2022"), "children_from_ukr_temp_prot_eu.R")

tableboom::inspect_r(path)

HTML table returned by tableboom::inspect_r()