PowerQuery Puzzle solved with R

#215–216

Puzzles

Author: ExcelBI

All files (xlsx with puzzle and R with solution) for each and every puzzle are available on my Github. Enjoy.

Puzzle #215

Today we have to do something that I really don’t like to do: constructing tables with esthetical structure, when esthethic is primarly for person reading and not having any knowledge bringing purpose. But today I found one purpose. After short transformation I decided to use gt package to do exactly what it have to do: make tables for people, not for machines. 🙂

Loading libraries and data

library(tidyverse)
library(readxl)
library(gt)

path = "Power Query/PQ_Challenge_215.xlsx"
input = read_excel(path, range = "A1:E20")
test  = read_excel(path, range = "G1:J15")

Transformation

result = input %>%
  mutate(out_day = case_when(
    !is.na(`Paid Date`) ~ NA_real_,
    `Due Date` > today() ~ 0,
    TRUE ~ as.numeric(difftime(today(), `Due Date`, units = "days"))
  )) %>%
  filter(!is.na(out_day)) %>%
  arrange(`Branch ID`, Customer, `Due Date`) %>%
  select(-`Paid Date`) %>%
  group_by(`Branch ID`)  %>%
  gt() %>%
  # change column names
  cols_label(Customer = "Branch ID / Customer",
             `Due Date` = "Due Date",
            `Loan Amt` = "Total Loan Amount",
             out_day = "Total Outstanding Days")

Presentation

Puzzle #216

That was pretty nice and brain warming puzzle unless I shouted Eureka! First look and I realized that have some pivoting and mutating, just normal job for data enthusiast. And something come to my mind just few minutes later. Simple transposition — did the job.

Loading libraries and data

library(tidyverse)
library(readxl)

path = "Power Query/PQ_Challenge_216.xlsx"
input = read_excel(path, range = "A1:E6")
test  = read_excel(path, range = "A11:E16")

Transformation — pivoting etc.

result = input %>%
  pivot_longer(everything(), names_to = "Column", values_to = "Item") %>%
  mutate(Column = str_remove(Column, "Column"), 
         item_n = str_remove(Item, "Item") %>% as.numeric()) %>%
  arrange(Column) %>%
  mutate(rn = row_number(), .by = Column) %>%
  mutate(Column_label = paste0("Items ", min(item_n, na.rm = TRUE), " - ", max(item_n, na.rm = TRUE)), .by = rn) %>%
  select(Column_label, Item, Column) %>%
  pivot_wider(names_from = Column_label, values_from = Item) %>%
  select(-Column)

Transformation — transposing

result = t(input)
result = as.data.frame(result)
names(result) = names(test)

Validation

all.equal(result, test, check.attributes = FALSE)
#> [1] TRUE

Feel free to comment, share and contact me with advices, questions and your ideas how to improve anything. Contact me on Linkedin if you wish as well.

PowerQuery Puzzle solved with R was originally published in Numbers around us on Medium, where people are continuing the conversation by highlighting and responding to this story.