Slide 1 of 4

Physics through Computational Thinking

Micro Lecture: Fixed Points in one dimension dymanical systems
Ambar Jain
Dept. of Physics, IISER Bhopal

Outline
In this lecture you will learn about

1. fixed points in one-dimensional systems

2. linearization of one-dimensional systems about the fixed points

2 Micro Lec 01.nb

Load the Code

Slide 2 of
! Lets 4 by loading the code for 4th order Runge-Kutta, since we will be using this later in the lecture.
start
In[99]:= rk4F_, X0_, tf_, nMax_ := Moduleh, datalist, prev, rate1, rate2, rate3, rate4, next,
h = tf - X0〚1〛  nMax // N;
Fordatalist = {X0},
Lengthdatalist ≤ nMax,
AppendTodatalist, next,
prev = Lastdatalist;
rate1 = F @ prev;
h
rate2 = F @ prev + rate1 ;
2
h
rate3 = F @ prev + rate2 ;
2
rate4 = F @ (prev + h rate3);
h
next = prev + (rate1 + 2 rate2 + 2 rate3 + rate4);
6
;
Returndatalist;

 Micro Lec 01.nb 3

Fixed Points in 1-dimension

Slide 3 of 4 the first order ODE in one dimension
! Consider

x" = f (x) (1)

! A fixed point x0 is the point for which motion ceases, that is

x" &x0 = f (x0 ) = 0 (2)

! Consider for example

" - x3
x= +x (3)
3
! The plot below shows three fixed points of this problem
!
x or f(x)

stable

stable
unstable
x
0
- 3 3

! We can check this numerically, that is the solution flows to one of the two stable fixed points
4 Micro Lec 01.nb

- x3
In[132]:= rateFunc[{t_, x_}] = 1, + x;
3
initial = {0, 0.1};
data4 = rk4rateFunc, initial, 10, 50;
ShowListPlotdata4〚 ;; , 1 ;; 2〛, Joined → True, PlotMarkers → None, PlotRange → {- 4, 4},

Plot 3,- 3 , {t, 0, 10}, PlotStyle → Dashed

Out[135]=
2 4 6 8 10

-2

-4

! Consider initial conditions in all the four regions

 Micro Lec 01.nb 5

- x3
In[136]:= rateFunc[{t_, x_}] = 1, + x;
3
data1 = rk4[rateFunc, {0, - 0.01}, 10, 100];
data2 = rk4[rateFunc, {0, 0.5}, 10, 100];
data3 = rk4[rateFunc, {0, 4}, 10, 100];
data4 = rk4[rateFunc, {0, - 4}, 10, 100];
ShowListPlotdata1〚 ;; , 1 ;; 2〛, Joined → True, PlotMarkers → None, PlotRange → {- 4, 4}, PlotStyle → Magenta,
ListPlotdata2〚 ;; , 1 ;; 2〛, Joined → True, PlotMarkers → None, PlotRange → {- 4, 4}, PlotStyle → Blue,
ListPlotdata3〚 ;; , 1 ;; 2〛, Joined → True, PlotMarkers → None, PlotRange → {- 4, 4}, PlotStyle → Red,
ListPlotdata4〚 ;; , 1 ;; 2〛, Joined → True, PlotMarkers → None, PlotRange → {- 4, 4}, PlotStyle → Green,
Plot- 3, 3 , {t, 0, 10}, PlotStyle → Dashed, AxesLabel → {Style["t", 16, Bold, Black], Style["x", 16, Bold, Black]}

x
4

Out[141]=

2 4 6 8 10
t

-2

-4
6 Micro Lec 01.nb

Linearization about the fixed point - One dimension

Slide 4 of 4
! Linearizing our ODE around the fixed point provides little more clarity about the stability, further we can obtain an approximate solution in the
neighbourhood of the fixed point.
! Near the fixed point f (x) has the Taylor Expansion of the form

x" = f (x = x0 + δ x) = f (x0 ) + δ x f ′ (x0 ) + +δ x2 

" (4)
⇒ x = λ δx

! where λ = f ′ (x0 )
! Substituting this in the ODE, near x = x0 , gives a local linear system
"
δx = λ δx (5)

! which simply has a solution

δ x(t) = δ x0 ⅇλ t
(6)
⇒ x(t) = x0 + δ x0 ⅇλ t

! If λ = f ′ (x0 ) is positive then the solution is unstable (diverging away from the fixed point), while if λ < 0 then the solution is stable (converging into the
fixed point). Indeed, this is what we saw from the plot in the precious example.
λ>0 : unstable fixed point
λ<0 : stable fixed point (7)
λ = 0 : neutral fixed point (sometimes stable on one side and unstable on the other)

Example
! Consider for example

- x3
x" = +x (8)
3

! As analyzed earlier, we have fixed points at ± 3 and 0. Let’s calculate λ = f ′ (x0 ) at all these points:

f ′ - 3  = - x2 + 1 = -2 < 0 STABLE
x=- 3

(9)
 Micro Lec 01.nb 7

f ′ (0) = - x2 + 1x=0 = 1 > 0 UNSTABLE

f ′  3  = - x2 + 1 = -2 < 0 STABLE
x= 3

(9)