Talk:Wave function collapse: Difference between revisions

Line 3:

I've rebuilt the task description (and I threw in a couple example implementations, though some other algorithms would satisfy the task description). Does this seem clear enough or have I overlooked some important ambiguities? --[[User:Rdm|Rdm]] ([[User talk:Rdm|talk]]) 14:10, 11 July 2022 (UTC)

: Thanks, it is much better now. Maybe some explanation what is the adjacent, how sockets (suitable each other sides of blocks) are defined and overlapped by collapsed tiles. --[[User:Darkfrei|Darkfrei]] ([[User talk:Darkfrei|talk]]) 06:44, 12 July 2022 (UTC)

: I think the algorithm for solution should be moved to the talk page and the task described. The example as given can be summarized by letting " "=0 and "#"=1 and interpreting the tiles edges as binary numbers read clockwise:

Tile west north east south weight

1 0 0 0 0 12.8

2 2 2 2 0 12.8

3 0 2 2 2 12.8

4 2 0 2 2 12.8

5 2 2 0 2 12.8

I must fill an 8x8 grid called result using these tiles such that for n=0..6 g=0..6 east[result[n,g]]=west[result[n,g+1]] and south[result[n,g]]=[north[result[n+1,g]].

64xTile1 would be a valid result thus far, the references suggest the result should contain tiles in relation to the number of each used in the initial image. This can be achieved:

let n[g]=count of Tile[g] used in result - weight of Tile[g] in minimize sum of n[g]<sup>2</sup>--[[User:Nigel Galloway|Nigel Galloway]] ([[User talk:Nigel Galloway|talk]]) 12:58, 12 July 2022 (UTC)

== Vote for deletion ==