Exploring the gt Package

The goal and the inputs

Let’s imagine you are a hard working monster. You perform various tasks at locations across Monstropolis and keep track of your billing in a messy spreadsheet. If only there was an easy way to generate invoices for all your clients!

Alas, there is. First let’s create the messy invoice data.

# load up libraries

library(tidyverse)
library(gt)

# spreadsheet of work

messy_csv_job_count <- tibble(
  job = c("Cleaning Scare Floor", "Scaring Boo", "Punking Randal"),
  `000 Monster Ln` = sample(3, 3),
  `111 Monster Ln` = sample(3, 3),
  `222 Monster Ln` = sample(3, 3),
  `333 Monster Ln` = sample(3, 3)
)

# fee schedule for various jobs

fee_schedule <- tibble(
  job = messy_csv_job_count$job,
  fee = c(20, 40, 60)
)

# put it all togther

monster_fees <- left_join(messy_csv_job_count, fee_schedule) %>% 
  gather(address, count, `000 Monster Ln`:`333 Monster Ln`) %>% 
  mutate(`fees owed` = fee * count) %>% 
  group_by(address) %>% 
  nest()

The invoices, ie the output

We now have our cleaned up data. Let’s deploy gt to create wonderful-looking invoices. We can use the dreaded for loop to accomplish our task. Each iteration of the loop will pick out the address and create a separate gt table.

output <- list()

for(i in seq_along(monster_fees$address)) {
  
  # create a variable with the target address
  
  add_of_int <- monster_fees$address[i]
  
  # for each job at each address, create a data frame
  
  by_job <- monster_fees %>% 
    filter(address == add_of_int) %>% 
    unnest() %>% 
    select_all(str_to_title) %>% 
    select(-Address) %>% 
    select(Job, Count, Fee, `Fees Owed`)
  
  # add a row with totals and bind
  
  total_df <- tibble(
    Job = "Total",
    Count = sum(by_job$Count),
    Fee = weighted.mean(by_job$Fee, by_job$Count),
    `Fees Owed` = sum(by_job$`Fees Owed`)
  )
  
  address_df <- bind_rows(by_job, total_df)
  last_row <- nrow(address_df)  
  
  # and NOW THE GT, rendered in HTML to facilitate output
  
  address_df %>% 
    gt() %>% 
    tab_header(
      title = html(
        paste0("Invoice for Monster Services RendeRed", 
               "<br>", 
               "<u><b>", add_of_int, "</b></u>")),
      subtitle =  html('<span style="font-weight:bold;color:red">PAYMENT DUE DATE: DEC 31, 2020</span>')
      ) %>%
    fmt_currency(
      columns = vars(Fee, `Fees Owed`),
      decimals = 0,
      currency = "USD"
    ) %>% 
    tab_style(
      style = cells_styles(
        text_weight = "bold",
        bkgd_color = "white"),
      locations = cells_data(rows = last_row)
      ) %>% 
    tab_style(
      style = cells_styles(
        text_color = "white"),
     locations = cells_data(rows = last_row, columns = vars(Fee))
     ) %>% 
    tab_source_note(
      source_note = html(
        '<br>
        <div align="right">
        May there be no questions! <u>Send payments to</u>:
        <br>
        <br>
        Monster University
        <br>
        Attn: Sully
        <br>
        911 Boo Ln
        <br>
        Monstropolis 00000
        </div>')
    ) -> output[[i]]
}

Here is the first of our invoices…

…and the last of our invoices

And there you have it folks. A collection of beautiful invoices.

For the enterprising folks who need old fashioned pdf output, here is my recommendation. With the loop output saved create a chunk per job (e.g. output[[1]]), and insert <P style="page-break-before: always"> between. Knit to html, set your css, and save to PDF. Of course you could do it in latex, but this weasel opts for the easier html hack output!