Framsticks

Hi people,

Did anyone try to create a humanoid Fram ? It would be a real challenge to
do that...

-------------------------------------------------------------------
[A long time ago, in a galaxy far far away...]

i have big difference with velocity and distance.
for me velocity it's distance/time.
i suppose it's different for framstick and then how compute velocity.
What's mean wait4up ?

In basic quadrupede
MX[@*:-.4,2:-2.890,1 :-1.808](RX(FX,),MX[@-1
:1.849,0:3.094,0:-1.387][|-1 :1.287,0:.5](rX(,X),,RX(X,)),rX(,FX))
we have
[@-1 :1.849,0:3.094,0:-1.387]
why not
[@-1 :1.849,0:1.703] ?

alainbb@mailclub.net

A lifespan-directed experiment

Objective
I wanted to run a lifespan-directed environment. Lifespan depends on
energetic conditions; mainly on ability to find food. A general fitness
such as lifespan does not impose any restriction on creatures regarding how
to find food. I expected to see an improvement in the ability to move,
and ability to move in the direction of the food. I started with a creature
capable of simple swimming (I used water because it is easier to move).
I expected creatures will move faster, and be able to activey seek food
(using smell sensors). In this experiment none of these happened,
nonetheless I was able to identify some speciation and parasitism,
although quite rudimentary.

Conditions
50x50 world, water. Because I was interested in interactions, I set the
number of simulataneously simulated creatures to 4. Death is on.
For selection, only lifespan has non-zero weight (1). Genotype similarity
linear. 300 genotypes. Energetics: starting e. of stick: 100; e. of a
dead stick: 20; idle metabolism: 3/stick/100cyc; automatic feeding: 20;
ball energy; 250. f1 genotype.
(I wanted to enable destructive collisions, but I got Runtime errors.)

Initial genotype
RRFFX[ @G:0.662,* :-1.333]RFX[@G:0.802]MMMMX[|-1:-1.569][|-1:2.367,0:-0.294]rrSIXCXXrcfSI(RX)
A simple little swimmer, 7 sticks. It can swim at speed 0.0065. On average
it lives for 3992. If it does not find any food, it lives for 3333
(100/3*100). After eating one ball, it lives for 4523 ((7*100+250)/(7*3)*100).

Losing the ability to move
Instead of moving better, creatures lost their ability to move altogether.
The main reason for this is that moving around aimlessly is not more
advantageous than staying in one place. This seems paradoxical, but
the existence (appearance) of a food ball is equally likely in the current
position A and any other position B. This is because the placement of
the balls is completely random.
(Clarify when and where are balls placed.)

Other strategies
I observed two different 'strategies' for increasing lifespan without
moving. The first is being bigger, which means a higher probability of
touching a ball. The other is being smaller, which means lower
probability, but a higher lifespan in the lucky event of touching a ball,
because of reduced metabolic requirements. I observed both of these
strategies coexisting in the population.
As generations progressed, the big creatures evolved to be even bigger
(up to 80-100 sticks and more), which small ones stayed small.

Parasitism
Creatures leave an eatable corpse after dying. The amount of energy of
the corpse is proportional to the size, thus a bigger creature leaves more
energy. This benefits creatures which will live in the future -- all
of them. If creatures become big, they increase the overall energy
inflow of the system. This benefits other creatures, mainly other big
creatures, but also some small creatures. These small parasites have
some chance of hitting upon an energy ball, or a corpse of a big dead
creature. Observing the simulation shows that big creatures touch balls
frequently, but also big corpses (energy transfer from a previous
generation). Sometimes small creatures feed on a big corpse, in which
case they live for very long. I call this parasitism.

I ran about 4000 generations.
I have the genepool from every 100th gen.

.sim file and some resulting creatures below.

# created Fri Jul 21 13:22:23 2000
# by Framsticks simulator (30-May-00) [MS Windows]
sim_params:
model:1
oldneurons:0
capacity:300
delrule:0
descol:0
debug:0
AutoKill:1
cr_c:0
cr_life:1
cr_v:0
cr_gl:0
cr_nnsiz:0
cr_di:0
cr_vpos:0
cr_vvel:0
cr_norm:0
fitfun:0
fitm:2
fitma:5
enablestats:2
cr_simi:1
testvel:9
cr_energ:0
MaxCreated:4
p_nop:50
p_mut:40
p_xov:10
xov_mins:0
Energy0:100
grow:0
corpsen:20
e_meta:3
aging:0
em_stat:0
em_dyn:0
sun:0
feed:20
feede0:250
autosave:1
overwrite:0
filecomm:0
wrldtyp:0
wrldsiz:50
wrldmap:
wrldwat:8
wrldbnd:1
mut_str:5
mut_neu:10
mut_exmod:eE
mut_exrec:S
mut_exctl:
gm_repair:1
gm_xosegm:0
geno_f1_sm0:0.05
geno_f1_sm1:0.02
geno_f1_sm2:0.02
geno_f1_sm3:0.02
geno_f1_sm4:0.1
geno_f1_nm0:1
geno_f1_nm1:0.1
geno_f1_nm2:0.05
geno_f1_nm3:0.1
geno_f1_nm4:0.05
geno_f1_nm5:0.2
geno_f1_simNN:1
geno_f1_simSN:1
geno_f1_simSS:1
geno_f1_simNS:1
geno_f1_simStr:4
geno_f4_mut1add:0.2
geno_f4_mut1del:0.2
geno_f4_mutAdd2div:0.2
geno_f4_mutAdd2link:0.2
geno_f4_mutAdd2rep:0.1
genkonw0:1
genkonw1:0
genkonw2:0
genkonw3:0

Here are some genotypes from gen 4200

st_min_lifespan:3334
st_avg_lifespan:3915.52
st_max_lifespan:10714
st_min_velocity:0
st_avg_velocity:0.000100315
st_max_velocity:0.00327094
st_min_distance:0
st_avg_distance:0.219158
st_max_distance:2.64762

FmSIX[@ = :0.000]

(X[@ T :0.713,0 :1.676,T :-2.500])

rIX

(XXXXXRRRFFXMMXXIXMMSSXRRRFSXXXXXrXMMMiXSSIX[ 0:2.975][|-1:601.997]FXX[@G:-0.844]
XMXrSIXXaMMMXSSIX[ 0 :1.391][|*:1.898,1:-0.849]FXXXMXXaMMMMXXMXMMMMiXMSSIXXaMMMMX
XXIXMMSSXRRRFFSXXXXXrXMMMiXSSIX[ 0 :1.391][|-1:601.997,0 :-4.013]IXFX[@ 0 :0.059]
[ @G:-0.946,0 :-370.967]XRFXMMMMXSSIX[ @0 :479.988][|*:-0.294,1 :981.135],FXXXMXX
aMMMMqXXMXaMMMXSSIX[ 0 :1.391][|*:1.898,1 :-0.849]FXXXMXXaMMMMXXMXaFIX[@ T :2.686
]X[|T:2.896]MMXiXSSIX[ 0 :1.391][|-1:601.997,0:-1.562,1 :1.777]XSSIX[ 0 :1.391]XX
aFIX[@ T :-3.223]X[@T:0.802]RMMMMiXSSIX[ 0 :1.391]X[ 1 :-1.716]XIXXMMSSXRRRFFXX,X
XrXMMX[@0:-4.892]MMMMiXMSSIXXaMMMMXXIXMMSSXRRRFFSXXXXXrXMMMiXSSIX[ 0 :1.391][|-1:
1.365,1 :-0.660],FXX[@G:1.126]XMXrSIXXXaX)

(q(,,RRRlaaaCFFX),XSIXXXmXXXXX,qXfIX)XmsXCFIX

(XXWXFFXSSXRRRFFSXXXXXrXSSXRRRFFXXXXrXrFMSSIIXXX[ 0 :-4.192],XMqXXMXLaMMMXSSIX[ 0
:1.391][|*:1.898,1 :-0.849]FXXXMXXaMMMMXXMXaFIX[@ T :2.686]X[|T:2.896]MMXiXSSIX[
0 :1.391][|-1:601.997,0:-1.562,1 :1.777]aMMMMXXXIXMMSSXRRRFFSXXXXXrXX[ / :-0.254]
XaXXXMXXaMMMMqXXMXLaMMMXSSIX[ 0 :1.391][|*:1.898,1 :-0.849]FXXXMXXaMMMMXXMXaFIX[@
T :2.686]X[|T:2.896]MMXiXSSIX[ 0 :1.391][|-1:601.997,0:-1.562,1 :1.777]XSSIX[ / :
1.391][|*:-2.016,1 :-4.613]XFXX[@G:-1.772]XMXrSIXXaX)

I'm interesting by a speed quadrupede with same structure of the basic
quadrupede
10 stick, but more neuron or blend.
actually i have 0.016 for velocity

Does anyone know what the highest velocity reached by a fram is? And where
I might get its genotype so i can view it?
The fastest I've seen is Zac Elston's frog ~.045

thanks,

jw

Hello Everyone,

Just some more random observations from an amateur.

It seems that often the best path to evolving creatures for specific
fitness factors is to start with other, different fitness parameters,
evolve for a while, then change to the fitness parameters you really
want, and continue evolving.

For example, I started an experiment with a simple "X" genotype, and had
Distance as the primary parameter. I think I also had some very small
weights for a couple other parameters. After running for a few days on
a 700Mhz Pentium III, the best creatures were getting a distance of a
little over 300 in their 10000 cycle life span. A couple days later,
they hit 350, and seemed to have leveled off.

I then started a second experiment, also with an "X" genotype, and had
Vertical Position and Vertical Velocity as my primary parameters, with a
small value for Distance. After just running overnight, they were
getting distances of 350, even though the heaviest parameters were for
vertical position and vertical velocity, and horizontal distance had a
much smaller relative weight. I haven't tried it yet, but I believe
that if I now switched the parameters to have a higher distance weight
and continued evolution, I'd get creatures that did much greater
distance than those in my first experiment that started out with
Distance as the primary parameter.

I think that the evolution in the first experiment was getting pulled
toward a nearby attractor basin in the solution space, and settled at
the center of this attractor basin. It got "stuck" at this local
maximum, but there were other, better maxima elsewhere in the solution
space (I think this is a common problem in alife and GA work).

In the second experiment, the vertical velocity-oriented parameters
pulled it in a different direction in the solution space, which actually
led to a deeper attractor basin in the terms of the distance parameter.
It's almost as if the creature needed help "getting on it's feet," so it
could see a better place a little farther off. Kind of like helping
water up over a small rise, so it can make its way down to a deeper
place on the other side, deeper than it would have found on it's own.

Anyway, sorry for the long ramble. Just thought I'd share my thoughts
so far.

--Marty Rabens

I was playing with f4 genotype and here are:

tightrope dancer

/*4*/<<N[-2:-0.5]>X>X

and slightly modified marching creature

/*4*/<<N[G:5]>X><<N[G:5]>X>X

both manualy created, need some evolution for beter performance

They are not to fast and they can't move when laying upside down
but they almost 'walk'.

Mirek Sitek

I have a simulation parameter set I use for adjusting the NN without
adding or
removing legs. basically I create the structure, add in where I think
muscles and
neurons should be.. I set the morphology to 0 and all the morph settings
to 0. however
I still get legs added and I have to manually remove those genotypes
during the
evolution. This is especially troublesome when I am running the
simulation on my sgi
and I never see the process..just the end result.. Is there a way to
absolutely stop
framsticks from adding legs?

thanks
-zac

here is my no-morph.sim

# created Fri Jul 14 16:44:43 2000
# by Framsticks simulator (30-May-00) [MS Windows]
sim_params:
model:1
oldneurons:0
capacity:200
delrule:2
descol:0
debug:0
AutoKill:1
cr_c:0
cr_life:0
cr_v:1
cr_gl:0
cr_nnsiz:0
cr_di:1
cr_vpos:0
cr_vvel:0
cr_norm:1
fitfun:0
fitm:2
fitma:5
enablestats:2
cr_simi:0
testvel:9
cr_energ:0
MaxCreated:1
p_nop:33
p_mut:33
p_xov:33
xov_mins:0
Energy0:100
grow:0
corpsen:0
e_meta:1
aging:0
em_stat:0
em_dyn:0
sun:0
feed:0
feede0:10
autosave:20
overwrite:1
filecomm:0
wrldtyp:0
wrldsiz:50
wrldmap:
wrldwat:-1
wrldbnd:0
mut_str:0
mut_neu:10
mut_exmod:eE
mut_exrec:S
mut_exctl:
gm_repair:1
gm_xosegm:0
geno_f1_sm0:0
geno_f1_sm1:0
geno_f1_sm2:0
geno_f1_sm3:0
geno_f1_sm4:0
geno_f1_nm0:1
geno_f1_nm1:0.1
geno_f1_nm2:0.05
geno_f1_nm3:0.1
geno_f1_nm4:0.05
geno_f1_nm5:0.1
geno_f1_simNN:100
geno_f1_simSN:1
geno_f1_simSS:1
geno_f1_simNS:1
geno_f1_simStr:100
geno_f4_mut1add:0.2
geno_f4_mut1del:0.2
geno_f4_mutAdd2div:0.2
geno_f4_mutAdd2link:0.2
genkonw0:1
genkonw1:1
genkonw2:1
genkonw3:1

Here's a rolling blender critter. It's not real fast but it has cool way of
moving I think.

I haven't evolved it yet. I was just playing around trying to learn the
genotype encoding and got this:

X(X,RRMMX[@G:.5](X[@G:.5],X))

-jeff