1

Universitatea Babeş-Bolyai, Facultatea de Matematică şi Informatică

Secţia: Informatică engleză, Curs: Dynamical Systems

1. A brief introduction to Maple 8

 Each Maple command ends with a semicolon ( ; ) or a colon ( : ) and the computer does
not execute the command until the line is entered (by hitting the return key). So,
; or : must follow every command

 Maple has Help and the question mark ? inserted in the command line and followed by
the operator, the function, the command … leads to the help page of this. You are
strongly encouraged to use the on-line help, as it is emphasized also below.

 Operators ( ?+; )
addition: + subtraction: - multiplication: * Division: /
exponentiation: ^ or ** repetition: $

 Constants ( ?Pi; )
π: Pi ℮: exp(1) √ -1: I ∞: infinity

 Elementary functions ( ?exp; )

exponential: exp(x) natural logarithm: ln(x)
absolute value: abs(x) square root: sqrt(x)
trigonometric: sin(x), cos(x), tan(x), sec(x), cot(x), csc(x)
Note the trigonometric functions in Maple require angles measured in radians
inverse trigonometric functions: arcsin(x), arccos(x), arctan(x), arctan(y,x)
hyperbolic functions: sinh(x), cosh(x), tanh(x), sech(x), csch(x), coth(x)

 Assign a value to a variable using :=

Examples: a:=1; # here the variable is denoted ‘a’ and the assigned value is a number
expr:=x*y+2*x+y^3-5; #here the variable is denoted ‘expr’ and the
assigned value is an expression
eq:=x+2=3; #here the variable is denoted ‘eq’ and the assigned value is
an equation
Note that, to denote a variable one can use any letter or group of letters (word) except
the ones that already are assigned (for example Pi, exp, sin, D, …).

 Unassign Examples: a:=’a’; expr:=’expr’; eq:=’eq’;

Note that the command restart; unassign all of the variables and undo all the commands

 Evaluation of a number using floating point arithmetic

Examples: evalf(1/2); evalf(sqrt(3)); evalf(sqrt(3),15); evalf(Pi); evalf(Pi,20);

 The previously computed result is returned by % ( ?%; )

 2

 Useful commands when working with expressions ( ?expand; ?subs; ?eval )

expand factor simplify subs(x=1, expr) eval(expr, x=1)

 Define a function with a given expression

Examples: f:=x -> 2*x+1; # this defines a real function f of one variable x
g:=(x,y)->x*y+x^2-3; # this defines a real function g of 2 variables (x,y)

 For a defined function f one can compute its value in a given point, or the values of its
derivatives in a given point ( ?D; )
f(1): f(1); f’(1): D(f)(1); f’’(1): (D@@2)(f)(1);

Note that f(x); # will return the expression of f, while

D(f)(x); #will return the expression of the first order derivative of f, i.e. f’(x)
(D@@n)(f)(x); #will return the expression of the n-th order derivative of f

 Differentiation of a function with a given expression ( ?diff; )

Examples: diff(2*x+3,x); #will return 2 diff(2*x+3,t); will return 0
diff(f(x),x); #will return f’(x) diff(x(t),t); will return x’(t)
diff(f(x),x$n); #will return the n-th order derivative of f

 Integration of a function with a given expression ( ?int; )

Examples: int(cos(x),x); #will return sin(x)
int(cos(x),x=0..Pi); # will return 0

 Limit of a function with a given expression in a given point ( ?limit; )

Examples: limit( sin(x)/x, x = 0); #will return 1

Note that the commands diff, int, limit, expand, factor require expressions for arguments.
If an expression is assigned to a variable f (for example f:=2*x^3-3;) then only diff(f,x);
is correct. In this case do not write f(x)!
But if f is defined as function of variable x (for example f:=x->x^2-5;) then only
diff(f(x),x); is correct since f(x) is an expression this time, while f is a function.

 If a variable expr is defined as an expression, the command f:=unapply(expr,x); turns

the expression into a function f of variable x.

 Solve algebraic equations and systems with solve and fsolve

solve(equation,x); #will give the exact solutions, if possible, of a given equation
whose unknown is x. Maple may not list all solutions; for example,
when trigonometric functions are involved.
fsolve(equation,x); #will give an approximate solution found numerically
fsolve(equation,x=a..b); #it is of great help to indicate an interval [a,b] where to
look for the solution
solve({eq1,eq2},{x,y}); #will solve a system of two equations and two unknowns
denoted here x and y
 3

 Plot the graph of a real function of one or two real variables ( ?plot; )
Examples: plot(x^2+2*x-1, x=-5..6); #plots the graph of the function f(x)=x^2+2x-1
for x in the interval [-5,6].
Note the graph is a planar curve.
plot({sin(x),cos(x)}, x=-5..6); #plots simultaneously the graphs of the
functions f(x)=sin x and g(x)=cos x for x in the interval [-5,6].
plot3d(x^2+y^2, x=-2..2, y=-2..2); #plots in 3d the graph of the function
f(x,y)=x^2+y^2 for (x,y) in the rectangle [-2,2]x[-2,2].
Note the graph is a surface in 3d.

 Plot a planar curve given by parametric equations

Examples: plot([2-t^2, t-t^3, t=-2..2]); #plots the planar curve given by the
parametric equations x=2-t^2, y=t-t^3 when the parameter t varies in the interval [-2,2].
plot([ [2*exp(-t), exp(-2*t),t=0..3], [-2*exp(-t), -exp(-2*t),t=0..3] ]);
#plots simultaneously the planar curve given by the parametric equations x=2e^(-t),
y=e^(-2t), for t in the interval [0,3] and the planar curve given by the parametric
equations x=-2e^(-t), y=-e^(-2t), for t in the interval [0,3]

 Plot a planar curve given by an implicit equation ( ?implicitplot; )

This command works only within the plot package, hence one first needs to call it
using the command with(plots):
Examples: implicitplot(x^2+y^2=4,x=-3..3, y=-3..3); #plots a circle with the center
in the origin of coordinates and radius 2
implicitplot(4*x^2+y^2=4,x=-3..3, y=-3..3); #plots an ellipse
implicitplot(y=x^2, x=-3..3, y=0..10); #plots a parabola

 2D and 3D animation work only after calling the plot package, i.e. first we type
with(plots):
Once we execute the maple command, we need to left click on the picture.
At that point choose Animation and Play.
Examples: animate(exp(-(x + t)^2), x = -10..10, t = 0..5, frames = 50);
animate(sin(x*t),x=-4*Pi..4*Pi,t=0..1,color=red);
animate([r*cos(t), r*sin(t), t = 0..5], r = 0..1, frames = 50);
animate3d(cos(t*x)*sin(t*y), x = -Pi..Pi, y = -Pi..Pi, t = 1..2, color = cos(x*y));
animate3d(t*y^2-x^2+x^4,x=-2..2,y=-2..2,t=0..2);
 4

2. The command dsolve

Is used to find the general solution of a scalar differential equation or of a differential system.
Also, it can be used to find the solution of an IVP or of a problem formed by a differential
equation and any other type of conditions.

 First we show how to introduce a differential equation in Maple. Note that, first we have
to clarify the notations for the unknowns and, respectively, for the independent variable.
An important thing to remember is that, in Maple, we have to write each time the
independent variable.
Examples. The first order differential equation x'=x , where the unknown is the
function x of independent variable t, can be introduced in Maple as follows:
eq1:= diff(x(t),t) = x(t);
Here we chose to assign the differential equation to the variable eq1
The second order differential equation x''+5x'-6x=0 , where the
unknown is the function x of independent variable t, can be introduced in MAPLE as:
eq2:= diff(x(t),t$2)+5*diff(x(t),t)-6*x(t)=0;

 In order to find the general solution of each of the two equations we use dsolve
Examples. dsolve(eq1,x(t)); dsolve(eq2,x(t));
Notice that the general solution evidentiate as many arbitrary real constants _C1, _C2,
... as the order of the differential equation.

Notice that there is the possibility to know details on the method used by Maple to
solve the differential equation by setting infolevel[dsolve]:=3 . The default setting is 1.

 In order to obtain the solution of a problem consisting of one or more differential

equations and one or more conditions (like, for example, an IVP) we use again dsolve
Examples. To solve the IVP x’=x, x(0)=-3 , first we prefer to assign the initial
condition to another variable, ic1:= x(0)=-3; then to use the command
dsolve({eq1,ic1},x(t)); Maple will return x(t) = -3*exp(t)
To solve the IVP x''+5x'-6x=0, x(0)=0, x’(0)=-1 again we assign the
initial condition to a variable, i.e. ic2:= x(0)=0, D(x)(0)=-1;
and then we use the command dsolve({eq2,ic2},x(t));

To find the general solution of the system x’=-y, y’=x+y we type

dsolve({diff(x(t),t)=-y(t),diff(y(t),t)=x(t)+y(t)},{x(t),y(t)});
 5

 In the case that we need to work further with the expression of a solution, we need to
assign to a new variable this expression. We explain in the sequel how one can do this.
First it is convenient to assign to a variable the output of the command dsolve, i.e.
sol:=dsolve({diff(x(t),t)-x(t)=0,x(0)=3},x(t));
This output is the equality x(t) = -3*exp(t) , hence we do not have yet the
expression of the solution, which is the right-hand side of this equality.
One can obtain the expression using the rhs command sol1expr:= rhs(sol);
With this expression we can do all the operations explained in Lab1.
We explain the same in the case of an IVP for a system of 2 equations with 2 variables.
sol:=dsolve({diff(x(t),t)-x(t)=0,diff(y(t),t)+y(t)=0,
x(0)=3, y(0)=2},x(t));
The output is x(t) = -3*exp(t), y(t)=2*exp(-t)
One can obtain the expression of x(t) and, respectively y(t), using the rhs command as
follows
xsolexpr:= rhs(sol[1]); ysolexpr:= rhs(sol[2]);

3. The command piecewise

 Here we show how to define in Maple a piecewise-valued function. The command is
piecewise(cond_1, expr_1, cond_2, expr_2, ..., cond_n,
expr_n, expr_otherwise);
where expr_i is an expression, cond_i is a relation or Boolean combination of
inequalities, and expr_otherwise is an (optional) default expression.
The semantics are as in a case statement: if cond_1 is true then expr_1, else if
cond_2 is true then expr_2, and so on. expr_otherwise gives a default case which
corresponds to all conditions being false. The default for expr_otherwise is 0.
A condition can be a single equality or inequality, or a boolean combination of
inequalities, such as t<3 or 0<t and t<=Pi . Equalities cannot be used in a
boolean expression. The conditions can contain relations with polynomials, abs, signum,
or piecewise functions, such as 0<t^2-4 and 0<x or abs(x)<4 . In all cases, t
is assumed to be a real variable.
A function defined using piecewise can be differentiated, integrated, simplified,
plotted, and used in some types of differential equations.
To plot a discontinuous function we recommend to use the option discont=true .
 6

4. Linear Algebra
 First load some additional packages: with(Student[LinearAlgebra]):
with(LinearAlgebra): with(linalg):

 This is how we introduce vectors and matrices:

[1,5,0] this is a row vector
<1,5,0> this is a column vector
Matrix([[1,5,0],[-1,-5,0]]) this is the 2x3 matrix whose first
row is [1,5,0] and second row is [-1,-5,0].
Matrix(<<1,-1>|<5,-5>|<0,0>>) this is the same matrix, but it
is given by columns
DiagonalMatrix([1,2,3]) this is the 3x3 diagonal matrix with
the elements 1,2,3 on the main diagonal
IdentityMatrix(3) this is the 3x3 identity matrix
Once a matrix A is introduced, its elements are A[1,1], A[1,2] …

 This is how we compute the determinant, the inverse, the eigenvalues, the eigenvectors,
the characteristic polynomial of a square matrix A:
Determinant(A) or det(A)
inverse(A) or A^(-1)
Eigenvalues(A) the result is a column vector that contains the
eigenvalues of A
lam,P:=Eigenvectors(A) the result gives first a column vector that
contains the eigenvalues of A (the variable named lam contains this
vector) and a matrix (here named P) whose columns are linear
independent eigenvectors of A corresponding to the order of the
eigenvalues as in the vector lam.
CharacteristicPolynomial(A,r) we can specify the name of
the variable, here is r

 Multiplication of two matrices A and B is with a dot, instead of star: A.B

 This is how we compute the Jordan form of a square matrix A, i.e. the “simplest” matrix
similar to A. We have that A is diagonalizable if and only if its Jordan form is diagonal.
JordanForm(A) this is gives the Jordan form of A, denoted B
JordanForm(A, output=’Q’) this is gives the transition (or
similarity matrix), i.e. the invertible matrix P such that A=PBP^(-1). We
know that, when A is diagonalizable, the columns of P are linearly
independent eigenvectors of A.
 This is how we compute the exponential of a matrix A: MatrixExponential(A)
 This is how we compute the limit as t goes to infinity of exp(tA):
Map(limit,MatrixExponential(t*A),t=infinity)
 7

5. The command dfieldplot

 This command works only after loading the package DEtools. Hence, before we must
put with(DEtools):
 Given either a system of two first order autonomous differential equations, or a single
first order differential equation, dfieldplot produces a direction field plot. There can
be only one independent variable.
 The direction field presented consists of a grid of arrows tangential to solution curves.
For each grid point, the arrow centered at ( ) will have slope . For system of two

first order autonomous differential equations this slope is computed using / ,

where these two derivatives are specified in the first argument to dfieldplot.
 The command is dfieldplot(deqns, vars, trange, xrange, yrange)
where deqns is a list or set of first order ordinary differential equations, vars is a
list or set of dependent variables, trange is the range of the independent variable,
xrange is the range of the first dependent variable, yrange is the range of the
second dependent variable.
 Examples: Represent the direction field of y’(x)=y(x) in the box [-1,1]x[-2,2].
dfieldplot(diff(y(x),x)=y(x), y(x), x=-1..1, y=-2..2);
Represent the direction field of the autonomous system x’(t)=-y(t), y’(t)=x(t) in the box
[-1,1]x[-2,2].
dfieldplot([diff(x(t),t)=-y(t), diff(y(t),t)=x(t)],
[x(t),y(t)], t=-3..3, x=-1..1, y=-2..2);
 8

6. The command DEplot

 This command works only after loading the package DEtools. Hence, before we must
put with(DEtools):

 Given a set or list of initial conditions, and a system of first order differential equations or
a single higher order differential equation, DEplot plots solution curves, by numerical
methods. The default method of integration is method=rkf45.

 The command is DEplot(deqns, vars, trange, inits) where deqns is

a list or set of first order ordinary differential equations, or a single differential equation
of any order, vars is a list or set of dependent variables, trange is the range of the
independent variable, while inits is a set or list of lists; initial conditions for
solution curves
 Examples: Represent simultaneously the solution curves with initial values y(0)=1, and,
respectively, y(0)=-1 of the differential equation y’(x)=y(x).
DEplot(diff(y(x),x)=y(x), y(x), x=0..2,[[y(0)=1],[y(0)=-
1]]);

Represent simultaneously the solution curves with initial values y(0)=1, y’(0)=0 and,
respectively, y(0)=-1, y’(0)=0 of the differential equation y’’(x)=y(x).
DEplot(diff(y(x),x$2)=y(x), y(x), x=0..1,
[[y(0)=1,D(y)(0)=0],[y(0)=-1,D(y)(0)=0]]);

Represent the orbit corresponding to the initial values x(0)=1, y(0)=0 of the autonomous
system x’(t)=-y(t), y’(t)=x(t) .
DEplot([diff(x(t),t)=-y(t), diff(y(t),t)=x(t)],
[x(t),y(t)], t=0..9, [[x(0)=1,y(0)=0]]);

Represent simultaneously the orbits corresponding to the initial values x(0)=1, y(0)=0,
and, respectively x(0)=1, y(0)=1 of the autonomous system x’(t)=-y(t), y’(t)=x(t) .
 9

DEplot([diff(x(t),t)=-y(t), diff(y(t),t)=x(t)],
[x(t),y(t)], t=0..9, [[x(0)=1,y(0)=0],[x(0)=1,y(0)=1]],
linecolor=[green,gold]);

7. The command contourplot

 produces a set of level curves of the input function for a discrete set of values (i.e. levels)
of the third coordinate. It work only after loading the package plots: with(plots):
 Examples: Represent the level curves of H(x,y)=sin(xy) in the box [-1,1]x[-1,1]
contourplot(sin(x*y), x=-3..3, y=-3..3);

8. Finding the Jacobian matrix and its eigenvalues

We consider a function of two real variables with two real components denoted (f1(x,y), f2(x,y)).
In order to find the eigenvalues of its Jacobian matrix calculated in a point (p1,p2) we use the
commands Jacobian and eigenvalues. They work after loading the packages linalg and
VectorCalculus, hence first write with(linalg): with(VectorCalculus):
Then use

Jm:=Jacobian([f1(x,y), f2(x,y)],[x,y]);
A:=subs([x=p1, y=p2], Jm);
eigenvalues(A);

9. Map iterations
We show how to find the first 30 iterations of a function f starting from x0. Then we represent
them using the command pointplot.

> x:=x0;
> for i from 1 to 30 do x:=f(x): psi(i):=x: print(x); od:
> points:=[[n,psi(n)]$n=1..30]:with(plots): pointplot(points,
symbol=circle);
 10

10. Euler’s methods

We will apply Euler's method to solve the initial value problem y'=2xy+exp(x^2), y(0)=1, in the
interval [0,1.5] with step size h=0.1. Since we would like to compare with the exact values, first
we find the exact solution of this IVP and we represent the graph of the solution and the direction
field of the differential equation.

> restart:
> with(DEtools):
> f:=(x,y)->2*x*y+exp(x^2); this is the right-hand side of the differential equation
> dsolve({diff(y(x),x)=f(x,y(x)),y(0)=1});phi:=unapply(rhs(%),x);
> DEplot(diff(y(x),x)=f(x,y(x)),y(x),x=0.. 1.5, [[y(0)=1]],
y=1..25);
> h:=0.1; # this is the stepsize. Hence, the number of steps necessary to cover the interval
[0, 1.5] is 15
> x:=0; y:=1; # (0,1) is the starting point, as given by the initial condition
> for i from 1 to 15 do y:=y+h*f(x,y): psi(i):=y: x:=x+h:
print(x,y,phi(x),abs(y-phi(x))); od:
> points:=[[n,psi(n)]$n=1..15]:with(plots): pointplot(points,
symbol=point);plot(phi(t),y=1..1.5);

We will apply the now the improved Euler's numerical method. We have to change only the
numerical formula. We must restart!

> restart:
> phi:=x->(x+1)*exp(x^2);
> h:=0.1; x:=0; y:=1;
> f:=(x,y)->2*x*y+exp(x^2);
> for i from 1 to 15 do y:=y+h/2*f(x,y)+h/2*f(x+h,y+h*f(x,y)):
psi(i):=y: x:=x+h: print(x,y,phi(x),abs(y-phi(x))); od:
> points:=[[n,psi(n)]$n=1..15]:with(plots): pointplot(points,
style=point);plot(phi(t),t=1..1.5);