Version 0.01 of my fuze evolution script.
Stull very basic. When you start it'll prompt you for a start sequence. Enter it with space delimiters as: 1 2 3 4 5
The opcodes are 1=wiggle at 1.0 CI
2=wiggle at 0.8 CI
3=wiggle@0.6
4=wiggle@0.4
5=wiggle@0.2
6=noOp
7=Wiggle again
8=shake
9=mutate.
On every evolution a new program is generated with a 60% chance of replacement of an opcode and 20% for addition or removal, respectively. The new program is run and the better of old and new is used in the next iteration.
Note that this will NEVER do LWS. LWS before the end-game is bad.
Also, this script is in beta. Don't expect it to be good this version.
Please consider adding .03, .05, .07, and .9 to list of CI.
It couldn't get me out of my local minimum but when I restarted it exceeded my score and much more quite quickly.
Save best achieved by script and if it is stopped then retore to that best.
I can always use undo to restore to last position if I so choose but I can't restore recent best if it's too far away.
I actually overlooked that. I'll get a new version uploaded tonight.
New version: v2. Please note that I'll be updating this often as more testing and feedback happens.
Oops, I guess it overwrote the last one. Meh. Just delete the old one and grab the new one.
Function 15 "wiggle sidechains" doesn't work. Possibility to insert "shake" would be nice.
I did a lot of work to reposition myself to a score within 50 points of my absolutebest. Unfortunately the script rest to it.
Please save starting position as recentbest then restore to recentbest.
Also, when one cancels retore recentbest. Also restore ci.
It looks like I need to learn the new set of functions.
Script updated to use recent best instead. I honestly don't know how to make it restore recentbest when cancelled, nor can I find the documentation for this, so just manually restore recent best after killing it for now,