Advent of Code 2021

Day 8
R
Author

Liam D. Bailey

Published

December 13, 2021


The Data


Day 8 is really more a puzzle than a coding challenge. We’re given the inputs and outputs from a broken 7-segment display (like on a digital clock). We can imagine each segment of the clock as having a corresponding letter, from a (top) to g (bottom).

We can create a dictionary that will translate a string of letters into a corresponding number following the image above.

#Actual letter number combos
correct_codes <- 0:9
names(correct_codes) <- c("abcefg", "cf", "acdeg", "acdfg", "bcdf", "abdfg", "abdefg", "acf", "abcdefg", "abcdfg")

The only problem is that the data we are given doesn’t have the correct letters! Each row of the input data we are given includes all 10 possible numbers (0-9) but the letters used are wrong. We have to work out a number of rules to decode this information! See the full explanation for today’s challenge here.

#Load data
day8_data <- readr::read_delim(file = "./data/Day8.txt", delim = " | ",
                               col_names = c("inputs", "outputs"), show_col_types = FALSE)

head(day8_data)
# A tibble: 6 × 2
  inputs                                                     outputs            
  <chr>                                                      <chr>              
1 fdceba bafdgc abeg afbdgec gbeacd abced bgc fcdge bg bedgc bafdec cgefd gcebd…
2 gbfac fegbda fcedagb bea ea abcdef dgbfe gfabe dgea gbdfec gdea bgefdc bea ef…
3 eg dagef gbcfeda ageb cegbfd gfe dbefa facdg abfged cedbaf befda daefb egf gc…
4 edgacfb gcfd dgb degfab bcega bdagc cgafbd fbacd gd fceabd fbdac gd gdbcaf dgb
5 eaf bedgaf dbafc bfceag fedcbg eafdcgb debfa ae adge gdebf abcdfeg febdg ae d…
6 afbdc aefg ea edbacfg dbefg eab gcbfde abecgd bgefad bfdae gfea bfdea gbdef f…

Convert our data into nested lists so that it’s easier to work with.

day8_data <- day8_data %>% 
  mutate(inputs = stringr::str_split(inputs, pattern = " "),
         outputs = stringr::str_split(outputs, pattern = " "))


The Challenges


Challenge 1 & 2

In combination, the two challenges from Day 8 aim to decode the outputs so we’ll combine them together. The first rules we can use are simple. There are 4 numbers that can be easily identified just by the number of characters in the input.

  • 2 characters = 1
  • 3 characters = 7
  • 4 characters = 4
  • 7 characters = 8

So if we see an input ‘ab’ this must correspond to the number 1 or ‘cf’ when correctly translated to the 7-segment display. But we don’t know if ‘a’ is ‘c’ or ‘f’. We need some more rules to work this out. We’ll apply all these rules below:

numbers <- NULL
for (i in 1:nrow(day8_data)) {

  #Split input into list with individual letters
  split_input  <- stringr::str_extract_all(day8_data$inputs[[i]], pattern = "[a-z]{1}")
  split_output <- stringr::str_extract_all(day8_data$outputs[[i]], pattern = "[a-z]{1}")

  #If there are two values, it must correspond to c and f
  #Remove from the list afterwards
  number1 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 2))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 2))] <- NULL

  #If there are three values, it must correspond to a, c and f
  #Remove from the list afterwards
  number7 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 3))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 3))] <- NULL

  #If there are four values, it must correspond to b, c, d and f
  #Remove from the list afterwards
  number4 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 4))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 4))] <- NULL

  #If there are seven values, it must correspond to a-g
  #Remove from the list afterwards
  number8 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 7))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 7))] <- NULL

  #Letter that's not shared between 1 and 7 must be a
  a <- setdiff(number7, number1)

  #Value of 6 characters that contains the letters in 4 must be 9.
  #Remove from the list afterwards
  number9 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 6 & all(number4 %in% letters)))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 6 & all(number4 %in% letters)))] <- NULL
  
  #The missing letter in 9 is e
  e <- setdiff(number8, number9)

  #Remaining 6 letter character that includes the same letters as number1 must be 0.
  #Remove from the list afterwards
  number0 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 6 & all(number1 %in% letters)))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 6 & all(number1 %in% letters)))] <- NULL
  
  #The missing value here is d
  d <- setdiff(number8, number0)

  #Last remaining length 6 is 6
  number6 <- split_input[unlist(lapply(split_input, \(letters) length(letters) == 6))][[1]]
  split_input[unlist(lapply(split_input, \(letters) length(letters) == 6))] <- NULL
  
  #Missing value is c
  c <- setdiff(number8, number6)

  #The letter in 7 that wasn't just extracted is f
  f <- setdiff(number1, c)

  #Letter in 4 that hasn't yet been deciphered in b
  b <- setdiff(number4, c(c, d, f))

  #Remaining letter in 0 after removing all known ones is g
  g <- setdiff(number0, c(a, b, c, e, f))

  #Dictionary
  dict <- letters[1:7]
  names(dict) <- c(a, b, c, d, e, f, g)

  #Now we can decode our output
  output <- correct_codes[unlist(lapply(split_output, \(letters) paste(dict[letters][order(dict[letters])], collapse = "")))]

  #Paste all the decoded numbers together
  numbers <- append(numbers, as.integer(paste(output, collapse = "")))

}
numbers
  [1] 6233 4675 5572 5197 8513 4356  307 8507 4088 3756 9295 2569 5165 3061 2862
 [16] 1316 7879 2323 2593 5720 6695 7066 9013 9050 4501 2886 1326  734 9840 8003
 [31] 1752 4944 2038 9216 9250 4237 2145 7182 1101 5063 2136 8496 8523 3326 4598
 [46]  227 6537 3390 5065 9306 8460 6335 2664 6446 5550 4303 6955 3520 9376 2586
 [61] 6520 2121 6685 3109 2838 8257 3082 7739 7132 2770 5996  521 3470  940 2060
 [76] 6396  400 6050 2314 8464  262 9628 4719  706 8307 5355 4289 2422 2634  714
 [91] 2016 2778 6695  474 6970 3689 3305   12 5074 8060 5359 9331 8993 3307 9329
[106]  458 9872 7089 6058 5030 3634 5702 3756 4506 2957 8995 5657 8375  415 2412
[121] 3021 7241 8368 9298 6666 8567 2648 9621 3215 5533 4972 2236 8329 2582 2929
[136] 2089 3436 3594 1383 5555 5461 2268 6377 1440 9494 8314 9483 3267 3037 5547
[151] 9184 4900 6696 7304 1294 1349 8642 6126 7367 5525 3861 3393 2958 5312 2028
[166] 5058 8513 4258 1780 3937 5568 6236 5159  518 8059 1014 1080 8538   40 6722
[181] 3193 2519 5877 2460 3348 7329 8622 4380 3629  771 3734 6563 1900 4502 1472
[196] 6293 7392 8285 9928 7535

The answer of our final challenge is the sum of all these values.

sum(numbers)
[1] 982158


See previous solutions here: