Doing Advent of Code in G'MIC

Software G'MIC
1

I have always wanted to do Advent of Code, but this year is when I actually understood the prompt being given. So, hence, I decided to give it a go.

EDIT: Deleted github repository for now until I fix up a few things.

3 Likes
2

Day 5 looking really difficult to understand and do. Probably might not even be easily feasible in G’MIC with the input.

1 Like
3

Day 6 Answer released - Deleted (See first post)

I love how cursed it looks.

4

I finally have solved Day 5 Part 2. What a PITA. But, I’m proud regardless.

5

Seems like a lot of hard work. Not of all of them are as bad a fit for g’mic as I imagined (some would work in a grid quite well), but probably not many people could check your code!

7

I finished on Day 8. Finding the Part 2 answer with LCM leads to breaking. How come G’MIC doesn’t have lcm yet? Or multiple variable GCD or LCM?

p2_ans = 922657
p2_ans = 67353961
p2_ans = 3973883699
p2_ans = -91983485980753
p2_ans = 1.9135324588576046e+18

So, I resorted to Python to calculate lcm of my values within G’MIC.

I think I’ll stop with the AoC here.

Here’s my result:

image
image757×616 30 KB

EDIT:

Looked up an alternative to the standard LCM formula. This code works in G’MIC:

#@cli advent_of_code_day_8_answer:
#@cli : Code to Day 8 Advent of Code.
advent_of_code_day_8:
it nodes

# 1. Parse string as parsable images

split -,{_'\n'}
last_var_index={$!-2}
fill[0] i==_'R'?1
name[0] directions

A,Z={'AZ'}
1 store. Aimg
1 store. Zimg

foreach[^directions] {
 discard {_'\ '}
 split -,{_'='}
 {0,t}=$>
 char:=i[#0,h#0-1]
 if $char==$A
  $Aimg
  eval da_push(#-1,$>)
  store. Aimg
 fi
 if $char==$Z
  $Zimg
  eval da_push(#-1,$>)
  store. Zimg
 fi
 rm[0]
 crop 0,1,0,{h-2}
}

foreach[^directions] {
 split -,{_','}
 (${{0,t}}^${{1,t}})
 keep[-1]
}

append[^directions] x => nodes

# 2. Find answers

eval "
 const directions=$directions;
 const nodes=$nodes;
 const starting_node=$AAA;
 const node_to_reach=$ZZZ;
 const length_of_LR_seq=h#directions;
 p1_ans=pos=0;
 current_node=starting_node;
 while(current_node!=node_to_reach,
  current_node=(I[#nodes,current_node])[i[#0,pos++]];
  if(pos==length_of_LR_seq,pos=0;);
  ++p1_ans;
 );
 set('p1_ans',p1_ans);
 "

$Aimg $Zimg da_freeze[-2--1] 
append[-2--1] c

1,100%,1,1,:"begin(
  const directions=$directions;
  const nodes=$nodes;
  const length_of_LR_seq=whd#directions;
 );
 current_node=i0#-1;
 ending_nodes=crop(#-1,0,0,0,1,1,h#-1,1,1);
 steps=pos=0;
 while(find(ending_nodes,current_node,0,1)==-1,
  test_node=I[#nodes,current_node];
  current_node=test_node[i[#0,pos++]];
  if(pos==length_of_LR_seq,pos=0;);
  ++steps;
 );
 "

eval "
 const num_of_values=h#-1;
 lcm(a,b)=(
  if(a>b
  ,A=a;B=b;
  ,A=b;B=a;
  );
  lcm=A;
  while(lcm%B,lcm+=A;);
  lcm;
 );
 p2_ans=i[#-1,0];
 for(p=1,p<num_of_values,++p,
  p2_ans=lcm(p2_ans,i[#-1,p]);
 );
 set('p2_ans',p2_ans);
 "

e[^-1] "Number of steps from AAA to ZZZ: "$p1_ans
e[^-1] "Number of steps for ghosts: "$p2_ans

rm
8

Someone needs to make a request, code or talk about it, which is what I do all the time: giving nudges and prods. Yeah, giving an example usually helps. Then, hopefully, David swoops in with a better implementation.