Day 14 Notebook

input<-read_lines("Day14Sample.txt")

robotframe<-as.data.frame(matrix(ncol=4,nrow=0))
for(i in 1:length(input)){
  y<-as.numeric(unlist(str_extract_all(input[i],"-?\\d+")))
  robotframe<-rbind(robotframe,y)}
colnames(robotframe)<-c("px","py","vx","vy")

Part 1

Move everything about in the frame and then count how many are in each quadrant

safetyfactor<-function(rf,max_x,max_y,s){
  midx<-max_x%/%2
  midy<-max_y%/%2
  rf<-rf%>%rowwise%>%
    ### find the final x & y position
    mutate(finalx=((px+s*vx)%%max_x),finaly=((py+s*vy)%%max_y)) %>%
    ### find which quadrant it is in
    mutate(quad=if_else(finalx<midx&&finaly<midy,1,
                        if_else(finalx>midx&&finaly<midy,2,
                                if_else(finalx<midx&&finaly>midy,3,
                                        if_else(finalx>midx&&finaly>midy,4,0)))))
  sf<-rf%>%count(quad)%>%filter(quad!=0)%>%arrange(quad)
  sf<-sf$n
  sf}

p1<-safetyfactor(robotframe,11,7,100)
p1

[1] 1 3 4 1

part1<-prod(p1)
part1

[1] 12

Part 2

Because there is a repeat every 10403, animate and watch that long is a possibility, but that’s a lot of rendering and watching

My first idea was that the tree would be centered around the midline - so I was looking for symmetry across the quadrants - that went nowhere.

Then, I thought about looking for anomalies in the safety factor -

This slightly alters the safetyfactor function from above

safac2<-function(rf,max_x,max_y,s){
  midx<-max_x%/%2
  midy<-max_y%/%2
  rf<-rf%>%rowwise%>%
    ### find the final x & y position
    mutate(finalx=((px+s*vx)%%max_x),finaly=((py+s*vy)%%max_y)) %>%
    select(finalx,finaly)%>%
    ### find which quadrant it is in
    mutate(quad=if_else(finalx<midx&&finaly<midy,1,
                        if_else(finalx>midx&&finaly<midy,2,
                                if_else(finalx<midx&&finaly>midy,3,
                                        if_else(finalx>midx&&finaly>midy,4,0)))))
  sf<-rf%>%count(quad)%>%filter(quad!=0)%>%arrange(quad)
  sf<-sf$n
  sf<-c(sf,prod(sf))
  sf}

### find tree runs this through all repeats.  It also takes forever.

findtree<-function(rf,max_x,max_y){
  treeframe<-as.data.frame(matrix(ncol=6,nrow=0))
  for(i in 1:(max_x*max_y)){
    t<-safac2(rf,max_x,max_y,i)
    treeframe<-rbind(treeframe,c(i,t))}
  colnames(treeframe)<-c("dex","q1","q2","q3","q4","safetyfactor")
  treeframe}

Running this allowed me to look for clumps

quadsearch<-findtree(robotframe,101,103)

The plot shows that right around 7000, theres an especially large group of robots in q1.

And plotting that one specifically shows

plottree<-function(rf,max_x,max_y,s){
  midx<-max_x%/%2
  midy<-max_y%/%2
  rf<-rf%>%rowwise%>%
    ### find the final x & y position
    mutate(finalx=((px+s*vx)%%max_x),finaly=((py+s*vy)%%max_y)) %>%
    ### find which quadrant it is in
    select(finalx,finaly)
  rf}

sevenzerofivefive<-plottree(robotframe,101,103,7055)

second_7055<-ggplot(data=sevenzerofivefive,aes(x=finalx,y=finaly))+
  geom_point()+
  theme(axis.text.x = element_blank(),
        axis.ticks.x = element_blank(),
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.title.y = element_blank(),
        axis.title.x = element_blank(),
        panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        legend.position="none")+
  scale_y_reverse()+
  coord_fixed()
second_7055

Tree

Finally - here’s a sped up version of the 4 minutes before the tree appears.

Robots Wandering