Fleur Kelpin Dec 17, 2020
library(tidyverse)
n <- 8
steps <- 6
input <- readLines("day17.txt") %>%
str_split("") %>%
unlist() %>%
map_lgl(~ . == "#") %>%
matrix(nrow = n, byrow = T)
Starting with your given initial configuration, simulate six cycles. How many cubes are left in the active state after the sixth cycle?
dim <- c(
n + 2 * (steps + 1),
n + 2 * (steps + 1),
1 + 2 * (steps + 1)
)
state <- array(F, dim = dim)
state[7 + 1:n, 7 + 1:n, 8] <- input
ds <- expand.grid(dx = -1:1, dy = -1:1, dz = -1:1)
innerxy <- 2:(dim[[1]] - 1)
innerzw <- 2:(dim[[3]] - 1)
for (i in 1:6) {
pop <- reduce(1:nrow(ds),
function(arr, index) {
arr + state[
innerxy + ds$dx[[index]],
innerxy + ds$dy[[index]],
innerzw + ds$dz[[index]]
]
},
.init = array(0, dim = c(
2 * steps + n,
2 * steps + n,
2 * steps + 1
))
)
state[innerxy, innerxy, innerzw] <-
(!state[innerxy, innerxy, innerzw] & pop == 3) |
(state[innerxy, innerxy, innerzw] & between(pop, 3, 4))
}
sum(state)
## [1] 388
For some reason, your simulated results don’t match what the experimental energy source engineers expected. Apparently, the pocket dimension actually has four spatial dimensions, not three.
dim <- c(
n + 2 * (steps + 1),
n + 2 * (steps + 1),
1 + 2 * (steps + 1),
1 + 2 * (steps + 1)
)
state <- array(F, dim = dim)
state[7 + (1:n), 7 + (1:n), 8, 8] <- input
ds <- expand.grid(dx = -1:1, dy = -1:1, dz = -1:1, dw = -1:1)
for (i in 1:6) {
pop <- reduce(1:nrow(ds),
function(arr, index) {
arr + state[
innerxy + ds$dx[[index]],
innerxy + ds$dy[[index]],
innerzw + ds$dz[[index]],
innerzw + ds$dw[[index]]
]
},
.init = array(0, dim = c(
2 * steps + n,
2 * steps + n,
2 * steps + 1,
2 * steps + 1
))
)
state[innerxy, innerxy, innerzw, innerzw] <-
(!state[innerxy, innerxy, innerzw, innerzw] & pop == 3) |
(state[innerxy, innerxy, innerzw, innerzw] & between(pop, 3, 4))
}
sum(state)
## [1] 2280