Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata...

30
Lectures on Lectures on Lectures on Lectures on Cellular Automata Cellular Automata Cellular Automata Cellular Automata Continued Continued Continued Continued Modified and upgraded slides of Martijn Schut [email protected] Vrij Universiteit Amsterdam Lubomir Ivanov Department of Computer Science Iona College and anonymous from Internet

Transcript of Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata...

Page 1: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Lectures onLectures onLectures onLectures onCellular Automata Cellular Automata Cellular Automata Cellular Automata ContinuedContinuedContinuedContinued

Modified and upgraded slides of

Martijn [email protected] Universiteit Amsterdam

Lubomir IvanovDepartment of Computer ScienceIona College

and anonymous from Internet

Page 2: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

MorphogeneticMorphogeneticmodelingmodeling

Dynamical SystemsDynamical Systemsand Cellular Automataand Cellular Automata

Page 3: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

matterheat

matterheat

matterenergymatterenergy

entropy dissipator

Organisms aredissipative processes

in space and time

Page 4: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Dynamical system

Time

Space

8

S(�8

x , t)

8

state S at�8

x at time t ?

Energy

Page 5: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Which space? - which geometry?

parameter space

a

b

geometric space

x

y

x2

a2 +y2

b2 =1

state space

x =a costy = b sin t

morphometric space

Q=S/A

S

A

?

time

velocity

θ

Page 6: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Which space? - which geometry?

? parameter space

genotypeenvironment

state space

phylogeny/ontogeny

morphometric space

morphology

geometric space

ontogeny/phenotype

Page 7: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

• discrete time stepsT=0,1,2,3,…

• discrete cells atX,Y,Z=0,1,2,3,...

• discretize cell statesS=0,1,2,3,...

Discretization has many aspects

t

t+1

t+2We can mix discrete andcontinuous values in somemodels

Page 8: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Cellular AutomatonCellular Automaton

t

t+1

update rules

neighbours

S(i, t +1) = f (S(neighbours,t), update rules)

S(i,t)

S(i,t +1)

In standard CA the values of cells are discretized

Page 9: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

• Reduce dimensions: D=1, i.e. array of cells• Reduce # of cell states: binary, i.e. 0 or 1• Simplify interactions: nearest neighbours• Simplify update rules: deterministic, static

A simple Cellular Automaton

Page 10: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Effects of simplification

CA mo d el Mor ph oge n esi sS pa ti al d is creti za ti on ~ le ve l o f d e tai l

(organi sm , ce ll, mo lecule)T em pora l d isc reti za ti on ~ le ve l o f d e tai l

(cel l cycle , reac tion rate)R e duc ti on o f d im en sion → pro found e ff ec ts

(2D “G am e of Li fe” )B ina ri za ti on co m pu tati ona l conven ience

(01→ A , 10→ T, 00→ G , 11→ C)nea rest -ne ighbou rin terac ti ons

~ spa ti a l re st ric ti on s

sim p le upda te ru les;sim u lt ane it y , ub iqu it y

~ non- li nea rit y→ profound effe ct s

Page 11: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

ΣΣ

Wolfram's 1D binary CA

• cell array• binary states• 5 nearest neighbours• „sum-of-states“ update

rule:

• cell array• binary states• 5 nearest neighbours• „sum-of-states“ update

rule:

S(i, t +1) = f S(i + j, t)j= −2

j =2

Wolfram, S., Physica 10D (1984), 1-35

S(i, t)S(i −1, t) S(i +1, t)

S(i, t +1)

S(i − 2, t) S(i + 2, t)

Observe importance ofsymmetry of logic function

Page 12: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

ΣΣ

CA rules

S ∈ {0,1}S ∈ {0,1}

Σ ∈ {1,2,3,4,5}Σ ∈ {1,2,3,4,5}

S

→ 25=32 different rules→ 25=32 different rules

Page 13: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Code for rule 6 (00110)

0S(i,T+1)

Σ neighbours 11

0

22

1

33

1

44

0

55

Page 14: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Rule codes0 0 0 0 0

0 0 0 0 1

0 0 0 1 0

0 0 0 1 1

0 0 1 0 0

0 0 1 0 1

1 1 1 1 1

rule 0rule 1rule 2rule 3rule 4rule 5

rule 32

Page 15: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Temporal evolutioninitial configuration (�t=0)

e.g. random 0/1

t=1

t=2

t=3

rule n

rule n

rule n

Page 16: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Temporal evolution

t=0

t=10

Page 17: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Rule 6: 00110

t=100

t=0

Page 18: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Rule 10: 01010

t=100

t=0

Page 19: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

CA: a metaphor for morphogenesis static binary

pattern

translated into state transition rule

rule iteration inconfigurationspace

genotype

translation,epigenetic rules

morphogenesisof the phenotype

00 11 00 11 00

0

11

1

22

0

33

1

44

0

55

Page 20: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

CA: a model for morphogenesis

update rule

iterative application of theupdate rule

CA state pattern

state space

genotype

epigenetic interpretation of thegenotype, morphogenesis

phenotype

morphospace

Page 21: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

CA rule mutations

11 00 0 1 0

0 1 0 1 11

swap

0 1 0 1 0

genotype phenotypedevelopmentnegate

How genotype mutations can change phenotypes?

Page 22: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

CA rule mutations0 1 0 1 0 1 0 0 1 0

0 1 1 0 0 1 0 1 0 0 1 0 1

0 1 0 0 1 0 0 0 1

0 1 1 1 0

1 1

0 1 0 1 0 1

negate

swaprule

Page 23: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Predictability?

pattern at t+∆t

pattern at t

explicit simulation:∆ iterations of rule n

?hypothetical"simple algorithm" ?

How to predict a next element in sequence? Tough!

Page 24: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Sum of naturalnumbers 1… N

algorithm 1:1+2+…+N

algorithm2:N(N+1)/2

(Gaussian formula)

Sum of naturalnumbers 1… N

algorithm 1:1+2+…+N

algorithm2:N(N+1)/2

(Gaussian formula)

Predictability

n1

N

Nth prime number

algorithm 1:trial and error

algorithm2: ?no general formula

Nth prime number

algorithm 1:trial and error

algorithm2: ?no general formula

pN ∈ pn{ }= p | p / i ∉ N;i ∈ 1, p] [{ }

easy Very tough!

Page 25: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

CA: no effective patternprediction

pattern at t+∆t

pattern at t

behaviour may be determined only by explicit simulation

Page 26: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

CA: deterministic, unpredictable,irreversible

• Simple rules generate complex spatio-temporal behavior

• For non-trivial rules, the spatio-temporalbehaviour is computable but not predictable

• The behavior of the system is irreversible• Similarity of rules does not imply similarity

of patterns

• Simple rules generate complex spatio-temporal behavior

• For non-trivial rules, the spatio-temporalbehaviour is computable but not predictable

• The behavior of the system is irreversible• Similarity of rules does not imply similarity

of patterns

Page 27: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Aspects of CA morphogenesisAspects of CA morphogenesis

• complex relationship between „genotype“and „phenotype

• effects of „genes“ are not localizable inspecific phenes (pleiotropy)

• phenes cannot be traced back to specificsingle genes (epistasisepistasis)

• phenetic effects of "mutations" are notpredictable

Page 28: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Meinhardt, H. (1995). TheAlgorithmic Beauty of Sea Shells.Berlin: Springer

Page 29: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Patterns and morphology• Pattern: a spatially

and/or temporallyordered distribution ofa physical or chemicalparameter

• Pattern formation

• Form (Size and Shape)

• Morphogenesis: Thespatiotemporalprocesses by which anorganism changes issize (growth) andshape (development)

Page 30: Cellular Automataweb.cecs.pdx.edu/~mperkows/temp/May22/B0028... · Lectures on Cellular Automata Continued Modified and upgraded slides of Martijn Schut schut@cs.vu.nl Vrij Universiteit

Where is the phene?

• Typification?• Comparative

measures?– length– density– fractal dimension

• Spatio-temporaldevelopment?

phenotype A

phenotype B