Genetic Turmit Evolver
python
c
Nevar pievienot vairāk kā 25 tēmas Tēmai ir jāsākas ar burtu vai ciparu, tā var saturēt domu zīmes ('-') un var būt līdz 35 simboliem gara.
Yann Weber bbc9ced95e Add a SWP swap operation (xchg two values on top of stack) 6 gadus atpakaļ
gte Add a SWP swap operation (xchg two values on top of stack) 6 gadus atpakaļ
tests Add a SWP swap operation (xchg two values on top of stack) 6 gadus atpakaļ
Doxyfile Corrected paths in Doxyfile 6 gadus atpakaļ
README Add README and runtest.sh script 6 gadus atpakaļ
README.txt Update README.txt 6 gadus atpakaļ
WOOT.result GTE is too smart and make diagonals changing direction every 50 steps... 6 gadus atpakaļ
runtest.sh Add README and runtest.sh script 6 gadus atpakaļ

README.txt

GTE Genetic Turmit Evolver

Use selection to evolve turmits. Turmit's behavior is defined by an expression
in reverse polish notation using an infinite loop stack and avoiding arithmetic
errors.


Exemple :
---------

# Evolve a pool of random generated turmit (with prog expr of size 5) the pool
# contains 15 turmits and only the best 1/5 is kept. The "best" is evaluated
# after 30000 steps and on an average of 3 runs. 2 threads will be used and
# a log with all expr and there fractdim will be written in gte.log
python3 -m gte evolve --threads 2 -L gte.log -p 15 -D 5 --steps 30000 -R 3 --prog-size 5

# Fast evolve from scratch
python3 -m gte evolve --threads 1 -L gte.log -p 16 -D 4 -R 5 -T 3 --steps 2000 -x 256 -y 256 -P '0' -E

# Generate an image of 40000 steps of 10 turmits in gte.png
python3 -m gte generate -o gte.png --steps 40000 -x 1024 -y 1024 -T 10 -P '0x982A POP 0xAE9D R & +'