Random numbers

 Numbers exhibit randomness

 They may be generated through some probabilistic randomness or through some strategical
methods
 They can be defined as numerical values taken by random variables
 Two Types
True random numbers Pseudo random numbers
 True random numbers:
 unpredictable
 generated in real physical processes
 radioactive decay, shaking of dice, tossing coin, Brownian motion, diffusion

 pseudo random numbers:

 generated artificially on computers
 are predictable
 repeated after certain period
• Randomness can be- Stochastic or Deterministic
• Stochastic: processes which vary randomly over various possible outcomes
• Deterministic: processes where randomness in output is determined in its input
 Random Number Generators
 True random number Generators
 Pseudo random Number Generators
 Pseudo Random Number Generators:
 arithmetic algorithms which when run on computers generate random numbers
 Applications: gaming, cryptography, computer simulation
 Require a mathematical Formula: recursive in nature
 Reproducible
 Periodic in nature
 Simulation allows the repetition in same simulation conditions
 May be defined over closed interval [0,1]
 require a seed value
Mid Square Generators
 Is the first algorithm used to generate pseudo-random numbers
 Have a four digit seed value
 new seed is obtained by squaring the input and chopping off at two ends
 A random number is obtained by dividing it by 10000
 Algorithm
Step1: choose total number of random numbers
Step2: give four digit seed value
Step3: start a loop over iterations
Step4:sqaure the seed: x=seed*seed
Step5: chop off the last two digits of x
Step6: chop off the first two digits of the number resulted in
step5
Step7: get the random numbers by dividing the obtained number
by 10000
Step8: print the number
 Step9: replace the seed by the number resulted in step6
  Step10:loop started at step3 ends
 Step11: program ends
Program
dimension x(500)
write(*,*)'no. of random numbers'
read(*,*)N
read(*,*)seed
do 50 i=1,N
x(i)=seed*seed
mx=x(i)/100
nx=mx/10000
seed=mx-nx*10000
x(i)=seed/10000
write(*,*)x(i)
50 continue
stop
end
Drawbacks
 Output is of poor quality
 Have short period
 Converges to zero or constant or cycles over short length

Multiplicative Congruential Method:

 Most commonly employed generator

 Known as method of power residues
 Xi+1 = (a Xi )mod m, where a ³ 0 and m ³ 0
 ‘a’ and ‘m’ are called magic numbers and are choosen to be large positive prime numbers
 Xi+1 is called residue and operation (a Xi) mod m
 Most natural choice for m is one that equals to the capacity of a computer word.
m = 2b (binary machine), where b is the wordlength in the computer word.
Period would be m/4
Algorithm

 Step1:Choose total number of random numbers

 Step2:Enter magic numbers a and m
 Step3:Take a seed x1
 Step4:Start a loop over i from 1 to N iterations
 Step5:Calculate the new number as Xi+1 = (a Xi )mod m
 Step6:Random number is obtained by dividing X i+1 by m
 Step7:Print the random number
 Step8:Loop at step4 ends
 Step9:Program ends

 C fortran program for multiplicative random number generator

dimension x(200)
open(3,file='multigen')
write(*,*)'no. of random numbers'
read(*,*)n
 read(*,*)a,m
read(*,*) seed
x(1)=seed/m
do 150 i=1,n,1
mx=x(i)
x(i)=x(i)-mx
write(*,*)x(i)
c write(3,*)x(i)
seed=x(i)*m
x(i+1)=(seed*a)/m

150 continue
stop
End
 Mixed Multiplicative Congruential Method:

An extension over multiplicative congruential method

Xi+1 = (a Xi +c)mod m, where a ³ 0 and m ³ 0

Tests for randomness

 Uniformity Test

 Auto-Correlation test

 Periodicity Test

 Serial auto Correlation Test

Uniformity Test

 Statstical data is divided into classes of equal sizes

 Bar charts are plotted to visualise whether each class has same no. of random numbers
 Numerical agreement is checked with chi-square test
 X2 =(1/no. Classes)* ∑ (observed-xpected)^2/(expected)^2
 for perfect distribution of random numbers X2 =0
Algorithm

 Step1:Give no. Of classes

 Step2:Give total no of random numbers
 Step3:Calculate expected number of random numbers in each class
 Step4:Find class size=1/no. Of classes
 Step5:Initialize the number in each class
 Step6: start a loop over N
 Step7: Start a loop over the classes count the no. Of random numbers in each class
 Step8: End loop started at step6
 Step9: end loop started at step5
 Step10:Calculate X2
 Step11: program ends
 Auto-correlation Test
 Tests the correlation between numbers and compares the sample correlation to the
expected correlation of zero.
 C(k)=
 C(k) => (i)th and (i+k)th particles are not correlated

Auto-Correlation test

 Algorithm
Step1: initialize sum for average and sqaure
Step2: start loop over random numbers
Step3: sum the numbers
Step4:sum the square of numbers
Step5: loop started at step2 ends
Step6: find mean of squares
Step7: give interval b/w two random numbers for which correln is to be checked
Step8: Initialize the sum of products of no.s in pairs,nl and no. Of pairs choosen for correln
Step9: calculate the pairs of random no.s
Step10:start loop over nl
Step11: calculate sum of product of random numbers
Step12:calculate sum of random numbers
Step13: end of loop at step 10
Step14: Find average of products
Step15: calculate c(k)
Step16: program end

Serial auto-correl’n test

 Modeled on radioactive decay of an array of nuclei placed in a square of L 2 lattice points

 than probability move forward and vice-versaAt t=0 all nuclei are assumed to be un-decayed
and represented by LxL (+) signs
 Co-ordinates will be formed of one of lattice site of nuclei which decay first
 Nucleus are allowed to be decayed one by one and repeated for L 2 times
 Count the no. Of nuclei left un-decayed
 Graph is plotted b/w left undecayed nuclei and no. Of trials
 Exponential decayà no. Correlations
otherwise-àstrong correlations

Periodicity Test

 Test is performed to check the length of period of generator

 Graphical representations: displacement vs no. Of steps
 If random no. Is less
Monte-Carlo method
 The Monte Carlo method is a numerical method
 for statistical simulation which utilizes sequences of random numbers to perform
the simulation
 Hit and miss method
 Sample mean method
 Hit and Miss method

 enclose the irregular shape into regular shape

 shoot the two figures blind foldedly

 probabiltiy of hitting is

Find the value of PI

The area of square=(2r)²

The area of circle = r²

area⋅of⋅square 4 r 2 4
= =
area⋅of⋅circle πr 2 π

area⋅of⋅circle
π =4∗
area⋅of⋅square

area .of .circle ¿.of .dots .inside .circle

=
area.of .square total.number .of .dots

Hit and miss algorithm

Generate two sequences of random numbers
Start from s=zero
Yj=Ri and X=Rj
If (X²+Y²<1) s=s+1
# of dots inside circle=s
total number of dots=N
π =4∗S / N
 fortran program for pi

data j,k,l/1000,53,1/
data j1,k1,l1/1000,53,1/
 write(*,*)'no. of hits'
read(*,*)n
count=0
do 110 i=1,n
l=mod(k*l,j)
rn=float(l)/float(j)
l1=mod(k1*l1,j1)
rn1=float(l1)/float(j1)
write(*,*)rn,rn1
if((rn*rn+rn1*rn1).le. 1)count=count+1
write(*,*)count
110 continue
integ=count
pi=4*integ
write(*,*)pi
stop
end
Monte Carlo Integration for multi-dimensional integrals

 Standard deviation

program for double integral

f(x,y)=x*x*y
write(*,*)'enter limits'
read(*,*)a1,b1,a2,b2
write(*,*)'enter random numbers'
read(*,*)n
 sum=0.0
data j,k,l/1000,53,1/
data j1,k1,l1/1000,53,1/
do 10 i=1,n
l=mod(k*l,j)
rn=float(l)/float(j)
l1=mod(k1*l1,j1)
rn1=*float(l1)/float(j1)

sum=sum+f(rn,rn1)
10 continue
sum=(b1-a1)*(b2-a2)*sum/n
write(*,*)sum
stop
end

Monte-Carlo simulation of radioactive decay

 Spontaneous decay of nuclei accompanied by the emmission of one or more particles

 Changes in atomic number and atomic mass occurs
 Alpha decay
 Beta decay
 gamma decay
 212Po84 -----> 208Pb82 + 4He2
 14O8 -----> 14N7 + e++ve
 24Mg12 -----> 24Mg12 + γ
 Modeled on basis of probability of a nuclei to decay
random no.=x ; prob. = p; parent nuclei=N;
daughter nuclei =p
X<p N=N-1 and D=D+1
C Program for radioactive decay of nuclei
write(*,*)'enter the values of N0,itmax'
read(*,*)No,itmax
N=No
Ido=0
P=0.25
IR=46
write(*,*)'Time No Do'
Do 60 IT=0,itmax
write(*,7)IT,N,Ido
nud=N
do 40 j=1,nud
call rand(ir,x)
if(x.lt.p)then
N=n-1
ido=ido+1
else
endif
40 continue
if(n .le. 0) goto 12
60 continue
7000
6000
5000
4000
No
3000
Do
2000
1000
0
7 format(3x,i3,2i10)
12 stop
end
subroutine rand(ir,x)
data m,k/1234,532/
multi=k*IR
IR=mod(multi,m)
R=float(ir)/float(m)
x=r
return
End
 RK-method for
solving 2nd order
differential
equation

Simulation of RC circuit:Charging of condenser

Discharging of condenser
 fortran program for rc-circuit
f(q)=v/r-q/(r*c)
open(3,file='rc')
v=4.0
r=2.0
c=1.0
h=1.0
q0=0.0
t0=0.0
qfinal=q0
i=0
write(*,*)' i time charge'
write(3,*)'i time charge'

write(3,*)i,t0,q0
7 i=i+1
 s0=f(q0)
s1=f(q0+s0*h)
q1=q0+h*(s0+s1)/2.0
t0=t0+h
write(3,*)i,t0,q1
if(abs(qfinal-q1) .ge. 0.0001)then
q0=q1
qfinal=q1
go to 7
endif
stop
end
LCR circuits

<= driven LCR circuit

 Program

C growth of current in LCR circuit

f(t,q,ci)=(-1.0*q)/(al*c)-(r/al)*ci
c=1.0
r=0.2
al=1.0
q0=0.0
h=0.1
ci0=1.0
t0=0.0
z0=ci0
tmax=15.0
i1=0
write(*,*)'n time current'
write(*,*)i1,t0,ci0
7 i1=i1+1
s0=z0
p0=f(t0,q0,z0)
p1=f(t0+h,q0+s0*h,z0+p0*h)
z1=z0+h*(p0+p1)/2.0
s1=z1
q1=q0+h*(s0+s1)/2.0
write(*,*)i1,t0,q1,z1
t0=t0+h
if(t0 .le. tmax) then
z0=z1
q0=q1
goto 7
endif
stop
End
LCR
 1.5

0.5
charge
current
0
0 2 4 6 8 10 12 14 16

-0.5

-1

LCR-driven

1.5

0.5

0 charge
0 2 4 6 8 10 12 14 16 charge-driven
-0.5

-1

-1.5

-2

-2.5

Hydrogen atom-eigen value problem

c program for sch equ hydrogen atom

pot(r)=(-1.0/r)+(l*(l+1)/(2.0*r*r))
read(*,*)l,rf,ro
read(*,*)e,h
i=0
r=0.05
zo=-1.902459
write(*,*)i,r,ro,rf
50 i=i+1
so=zo
po=-2.0*zo/r-2.0*(e-pot(r))*ro
p1=-2.0*(zo+po*h)/(r+h)-2.0*(e-pot(r+h))*(ro+so*h)
z1=zo+(po+p1)*h/2.0
s1=z1
r1=ro+(so+s1)*h/2.0
r=r+h
ra=2.0*exp(-r)
write(*,*)i,r,r1,ra
zo=z1
ro=r1
if(r.le.rf) goto 50
stop
end

R(r)
2
R(r)
1

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Interference pattern
Algorithm

Step1: Define

Step2: Input parameters Λ, A D, d, N Ymax,T, NPTS

Step3: calculate h=Ymax/NPTS

Step4: start loop over points to calculate intensity

Step5: Find r1 and r2

Step6: intialize intensity I=0

Step7: Begin loop over time steps N in one cycle

Step8: calculate resultant amplitude A= f(r1,r2,t)

Step9: sum the intensity I=I+Amplitude 2

Step10: end loop started at step7

Step11: find average intensity = I/N

Step12: loop begun at step 4 ends

Step13: Program ends

 c program for interference

real k,ints
common a,d,k,nt
open(3,file='inter')
write(3,*)'enter the no. of slits ns'
read(*,*)ns
write(*,*)ns
write(*,*)'enter distance between two slits'
read(*,*)a
write(*,*)'enter the distance of D'
read(*,*)D
write(*,*)'enter the wavelength'
read(*,*)lam
write(*,*)'max position'
read(*,*)tmax
write(*,*)'enter no of points on the screen'
read(*,*)nt
alam=lam*1.0e-7
pi=3.14159
k=2.0*pi/alam
dy=ymax/npoint
write(3,*)' position
intensity'

do 50 ipoint=0,npoint
y=ipoint+dy
call swave(y,ints)
write(3,*)y,ints
50 continue
continue
stop
end
subroutine swave(y,ints)
real k,ints
integer t
common a,d,k,nt
fun(r,tx)=amod((k*r-omg*tx),pi2)
pi=3.14159
pi2=2.0*pi
omg=2.0*pi/(2.0*nt)
ints=0
do 20 t=1,nt
tx=t
y1=y-0.5*a
 y2=y-0.5*a
r1=d*d+y1*y1
r2=d*d+y2*y2
r1=sqrt(r1)
r2=sqrt(r2)
sum=sin(fun(r1,tx))/r1+sin(fun(r2,tx))/r2
ints=ints+sum*sum
20 continue
ints=ints/nt
return
end
Diffraction Pattern
 2dsin=nλ

Algorithm

Step1: define
Step2: input parameters: λ, A, Ymax, D, Npts,no.of time steps N, size of grating
Step3: Find wave number,k and angular frequency ω
Step4: Find step size h=Ymax/Npts
Step5: begin loop to find intensity over screen points
Step6: find distance of the slits from points on the screen
Step7: initialize intensity I= 0
 Step8: begin loop over time steps N
Step9:initialize sum of amplitudes, sum=0
Step10: begin loop over number of lines
Step11: find wave amplitude and add to sum
Step12: loop begun sat step 13 ends
Step13: intensity I= I + total Amplitude 2
Step14: end of loop begun at step 11
Step15: find average intensity = I/tN
Step16: loop begun at step 7 ends
Step17: program ends
 Fortran Program
c program for diffraction
real k,ints
common ns,a,d,k,nt
open(3,file='diffr')
write(*,*)'enter the no. of slits'
read(*,*)ns
write(*,*)'enter the distance bw two slits'
read(*,*)a
write(*,*)'enter the distance D'
read(*,*)d
write(*,*)'enter the wavelength'
read(*,*)lam
write(*,*)'enter maximum position of ymax in mm'
read(*,*)ymax
write(*,*)'enter no of points on the screen npt'
read(*,*)npt
write(*,*)'enter the time period'
read(*,*)nt
alam=lam*1.0e-7
pi=3.14159
k=2*pi/alam
dy=ymax/npt
write(3,*)'position intensity'
do 50 ipt=-npt,npt
y=ipt+dy
call swave(y,ints)
write(3,*)y,ints
50 continue
stop
end
subroutine swave(y,ints)
real k,ints
integer t
common ns,a,d,k,nt
fun(r,tx)=amod((k*r-omg*tx),pi2)
pi=3.14159
pi2=2.0*pi
del=a/(ns-1)
omg=2*pi/(20.0*nt)
ints=0
 do 20 t=1,nt
tx=t
sum=0.0
do 30 is=1,ns
y1=-0.5*a+(is-1)*del
r1=d*d+(y-y1)*(y-y1)
r=sqrt(r1)
sum=sum+cos(fun(r,tx))/r
30 continue
ints=ints+sum*sum
20 continue
ints=ints/nt
return
end
 Oscillatory Motion
  Program for simple harmonic oscillator

C program for SHO

dimension t(2000),A(2000),v(2000),s(2000)
write(*,*)'k,m'
read(*,*)k,M
write(*,*)'xo vo tmax h'
read(*,*)x0,v0,tmax,h
n=tmax/h
t(1)=0
s(1)=x0
v(1)=v0
do 230 i=1,n
t(i)=t(1)+h*i
a(i)=-1.0*k*s(i)/m
write(3,*)t(i),v(i),s(i),a(i)
c write(*,*)t(i)
v(i+1)=v(i)+a(i)*h
s(i+1)=s(i)+v(i)*h
230 continue
stop
end
Undamped

x(t)
2.00E+00
1.50E+00
1.00E+00
5.00E-01 x(t)
0.00E+00
0.00E+00
-5.00E-01 2.00E+00 4.00E+00 6.00E+00
-1.00E+00
-1.50E+00
-2.00E+00

Damped

1.2
1
0.8
0.6
0.4 x(t)
0.2 v(t)
0
-0.2 0 2 4 6 8 10 12 14 16

-0.4
-0.6
c driven harmonic oscillator
dimension t(2000),A(2000),v(2000),s(2000)
func(x,Y,Z)=(-k*x-c*Y+f0*cos(omega*Z)/m)
read(*,*)k,M,c
read(*,*)x0,v0,tmax,h
read(*,*)f0,omega
n=tmax/h
t(1)=0
s(1)=x0
v(1)=v0
a(1)=func(s(1),v(1),t(1))
do 230 i=1,n
t(i)=t(1)+h*i
write(3,*)t(i),v(i),s(i),a(i)

s(i+1)=s(i)+v(i)*h
a(i+1)=func(s(i+1),v(i),t(i+1))
v(i+1)=v(i)+a(i+1)*h

230 continue
stop
end

 Simple pendulum
 program for simple pendulum
c program for simple pendulum
 write(*,*)'h,tf'
read(*,*)h,tf
g=9.8
l=1
tho=0.25
to=0.0
zo=0.0
i=0
write(*,*)to,tho,zo
50 i=i+1
so=zo

po=-g*sin(tho)/l
p1=-g*sin(tho+so+h)/l
z1=zo+h*(po+p1)/2.0
s1=z1
th1=tho+h*(so+s1)/2.0
to=to+h
write(*,*)to,th1,z1
zo=z1
tho=th1
if(to .lt. tf) goto 50
stop
end
c program for chaotic pendulum
f(x,v,t)=-1.0*c*v-(1+a*cos(w*t))*sin(x)
open(3,file='chpend')
write(*,*)'tmax,h'
read(*,*)tmax,h
write(*,*)'xo vo'

read(*,*)xo,vo
write(*,*)'c a w'
read(*,*)c,a,w
i=0
x=xo
v=vo
50 i=i+H
so=vo
k1=f(x,v,t)
k2=f(x+h*so/2,v+k1/2,t+h/2)
s1=vo+k1*h/2
k3=f(x+s1*h/2,v+k2*h/2,t+h/2)
s2=vo+k2*h/2
k4=h*f(x+s2*h,v+k3*h,t+h)
v=v+((h/6)*(k1+2.0*k2+2.0*k3+k4))
s3=v
x=x+((h/6)*(s1+2.0*s2+2.0*s3+s4))
write(*,*)t,x,v
 t=t+h
if(t .lt. tmax) goto 50
10 continue
stop
end
 Phase trajectories of chaotic pendulum

F(x,v,t)=-cv-(1+f cos(wt))*sin(x)
(i) c= 0 f=0

(ii) c= 1 f=0

(ii) c= 1 f=1

(iii) c= 1 f=3

(iv) c= 1 f=1.7
c growth of current in RL circuit
f(ci)=v/al-(r/al)*ci
v=4.0
r=2.0
al=1.0
h=0.1
ci0=0.0
t0=0.0
cifinal=ci0
i1=0
write(*,*)'n time current'
write(*,*)i1,t0,ci0
7 i1=i1+1
s0=f(ci0)
s1=f(ci0+s0*h)
ci1=ci0+h*(s0+s1)/2.0
t0=t0+h
write(*,*)i1,t0,ci1
if(abs(cifinal-ci1) .ge. 0.001) then
ci0=ci1
cifinal=ci1
goto 7
endif
stop
end

current
2.5

1.5 current

0.5

0
0 0.5 1 1.5 2 2.5 3 3.5
c

 growth of current in driven RL circuit

f(t,ci)=(v0*sin(omega*t))/al-(r/al)*ci
v0=1.0
omega=0.6
r=2.0
al=1.0
phi1=(-1.0*omega*al)/r
phi=atan(phi1)
cia0=v0/(sqrt(r*r+omega*omega*al*al))
h=0.1
ci0=0.0
 t0=0.0
cifinal=ci0
i1=0
write(*,*)'n time current'
write(*,*)i1,t0,ci0
7 i1=i1+1
s0=f(t0,ci0)
s1=f(t0+h,ci0+s0*h)
ci1=ci0+h*(s0+s1)/2.0
v=v0*sin(omega*t0)
t0=t0+h
cia=cia0*sin(omega*t0+phi)
write(*,*)i1,t0,ci1,cia,v
if(abs(cifinal-ci1) .ge. 0.001) then
ci0=ci1
cifinal=ci1
goto 7
endif
stop
end

1.2

0.8
current
0.6
current-ana
0.4 v

0.2

0
0 0.5 1 1.5 2 2.5 3 3.5

Chaos

 Chaos: Measurements on the system at given time might not allow the future to be
predicted with certainty, even if the force laws are known exactly! (Experimental
uncertainties in the initial conditions). Deterministic Chaos: Always associated with a system
nonlinearity. Nonlinearity: Necessary for chaos, but not sufficient! All chaotic systems are
nonlinear but not all nonlinear systems are chaotic.

 Chaos occurs when the system depends very sensitively on its previous state. Even a
tiny change in initial conditions can completely change the system motion.

 Chaotic Systems: Can only be solved numerically. Only with the availability of modern
computers has it become possible to study these phenomena. No simple, general rules for
when the system will be chaotic or not. Chaotic phenomena in the real world: Irregular
heartbeats, Planet motion in the solar system, Electrical circuits, Weather patterns, …
 Chaos- complete disorder/confusion/disorder
 Non-linearity in real physical processes
 Chaotic system appear stochastic but are deterministic and still unpredictable
 Chaotic systems are predictable at times and random at instants
 Is interinsic property
 Extremely sensitive to initial conditions (butterfly effect)
 Chaotic behavior can only be found in systems of equations with three or more variables.
 Attractors: set of points in phase space towards which system evolves

 Strange attractors are chaotic attractors

 Represent a stable structure of long term trjacteroies in a bounded region
 Trajectories fold back onto itself and result in divergence of nearby states
 Infinite no. Of thin layers having features of self-similarities: fractals
 Reduction in dimensionality implies only few degrees of freedom will be active
 Strange attractors are chaotic attractors
 Represent a stable structure of long term trjacteroies in a bounded region
 Trajectories fold back onto itself and result in divergence of nearby states
 Infinite no. Of thin layers having features of self-similarities: fractals
 Reduction in dimensionality implies only few degrees of freedom will be active
 Strange attractors are chaotic attractors
 Represent a stable structure of long term trjacteroies in a bounded region
 Trajectories fold back onto itself and result in divergence of nearby states
 Infinite no. Of thin layers having features of self-similarities: fractals
 Reduction in dimensionality implies only few degrees of freedom will be active
Poincare section

 These are snapshot of the motion in phase space taken at intervals

 Slices of phase space trajectories taken by time plane
 They are essentially 2d representations of 3d phase space diagram plots
 (x,v,t) space= (x,v,Tn) space  set of points (x,v,Tn)
 In the preesnce of driving forces t=2*pi/omega is taken as circumference of time circle of
poincare sections
 Informative as complex structures are consequence of folding of steady states
Bifurcation – Periodic doubling

 Bifurcation is the pint where a thresh hold stability is reached

 Achieved by increasing the value of control parameter to push farther from the system
 Near bifurcation fluctuations play role
 At this point symmetry is broken –leads to cascading of bifurcation associated with periodic
doubling
 Periodic doubling is route to chaos
Parameters to chaos

 Convergence parameter: ratio of control parameter in two successive cycles at bifurcation

 =4.9999

 Scaling number: : ratio of sets of scales of trajectories splitting in two successive cycles at
bifurcation

 = 2.50
 

Lyaponov exponents

 Quantitative measure of evolution of two nearby trajectories

 For initial states x and x+E
 Divergence is
 Lyp. Exponent gives stretching b/w two adjacent points
 Lyp exp. negative=> stable fixed point
 Lyp exp. Positive=>unstable fixed point
Logistic Mapping

 System represented by logistic map

 Logistic map is simplest generalisation
 Logistic equation
 xn lies in [0,1]
 With initial xo set of xi’s are generated à orbit
 Characteristics of logistic map
 Plot f(x) curve
 Plot diagonal line xn+1 =xn
 Plot vertical and horizontal lines as (xo,yo),(xo,y1) and so on
  features

 <1  fixed point at x=0

 1< <3  fixed point at x=1-

1/

 3< <3.5  periodic doubling

 >3.5  infinite periodic

doubling
 =0.8

=2.8

=3.45

=3.6