f(x) = e^x-3x matlab code

857 views (last 30 days)
Adomas Bazinys
Adomas Bazinys on 25 Feb 2018
Commented: Walter Roberson on 25 Oct 2022
I want to write a matlab script that finds solutions of function f(x) = e^x - 3x. I tried to write some code. My matlab code should include iteration table and graphic of function. Can anyone help me please with a code or give me advices?
f=@(x) exp(x) - 3*x ;
a = 0; b = 1;tol = 1e-8;
if (f(a)*f(b) > 0)
error ('invalid choice of interval')
end
r=0; n=0;
while ((b-a)/2 > 0)
n = n + 1;
r = (a+b)/2;
if (f(r) == 0)
break;
elseif (f(a)*f(r) <= 0)
b = r ;
else
a = r;
end
end
  3 Comments
Rik
Rik on 25 Feb 2018
Did you read Steven Lord's comment? Your current code finds 1 solution to this equation already.
If you have some restriction (e.g. because this is a homework assignment), please mention them.

Sign in to comment.

Accepted Answer

John BG
John BG on 25 Feb 2018
Hi Adomas Bazinys
What you want to do is, the way you want to do it, indeed good exercise, but
1.
What about fsolve
At Mathworks the function fsolve was time ago developed precisely so that the people using MATLAB could save time avoiding the need for any writing of such functions, and go straight to the roots, have a look:
fun = @(x) exp(x)-3*x;
x0 = [1.5,0];
x = fsolve(fun,x0)
x =
1.512134551657842 0.619061283355628
2.
Don't start with x0=[0 0] or [1 0] because then fsolve only returns a single real solution
3.
The requested graph
fplot(fun);grid on
.
4.
And the table
x_range1=0;x_range2=10;xrange_step=1;
x1=[x_range1:xrange_step:x_range2]';fx1=fun(x1);
T1=table(x1,fx1)
T1 =
11×2 table
x1 fx1
__ __________________
0 1
1 -0.281718171540954
2 1.38905609893065
3 11.0855369231877
4 42.5981500331442
5 133.413159102577
6 385.428793492735
7 1075.63315842846
8 2956.95798704173
9 8076.08392757538
10 21996.4657948067
I assumed you ask for a MATLAB table.
You can also export this table to Excel, Word or a simple .txt file, if you choose to call those the intended tables to hold the result.
5.
Checking the 4 limits +- Inf +-1j*Inf
syms n;limit(exp(n)-3*n, n, -inf)
=
Inf
syms n;limit(exp(n)-3*n, n, inf)
=
Inf
syms n;limit(exp(1j*n)-3*1j*n, n, inf)
=
NaN
syms n;limit(exp(1j*n)-3*1j*n, n, -inf)
=
NaN
Do you really need considering exp(1j*imx)-3*1j*imx ?
If so the start point is
exp(1j*imx)-3*1j*imx = cos(imx)+1j*(sin(imx)-3*imx)
this would be when real(x)>>imx=imag(x)
6.
the 2 real roots found with fsolve probably turn into a corona of complex roots that may be appreciated developing a bit further the following start point.
rangelim=15;rangestep=.01;
rex=[-rangelim:rangestep:rangelim];
imx=[-rangelim:rangestep:rangelim];
[RX,IX]=meshgrid(rex,imx);
Z=exp(RX+1j*IX)-3*(RX+1j*IX);
figure(2);hs=surf(abs(Z),RX,IX);hs.EdgeColor='none';
Adomas
If interested in the complex roots, please let me know.
If you find the so far supplied answer useful, would you please be so kind to consider marking my answer as Accepted Answer?
To any other reader, if you find this answer useful please consider clicking on the thumbs-up vote link
thanks in advance
John BG
  1 Comment
Steven Lord
Steven Lord on 25 Feb 2018
The fsolve function is part of Optimization Toolbox, but you don't need that toolbox for a one-variable, one-equation problem. The fzero function in MATLAB can handle those problems. Of course I suspect this is a homework assignment and the professor won't accept code that simply calls either fsolve or fzero.

Sign in to comment.

More Answers (2)

Rik
Rik on 25 Feb 2018
If you want to store intermediate result, us the n for indexing. Also, don't you mean if abs(f(r(n)))<=tol? Then you would actually use the variable tol.
f=@(x) exp(x) - 3*x ;
a = 0; b = 1;tol = 1e-8;
if (f(a)*f(b) > 0)
error ('invalid choice of interval')
end
r=0; n=0;
while ((b-a)/2 > 0)
n = n + 1;
r(n) = (a+b)/2;%#ok suppress warning, we can't know the length of r in advance
if (f(r(n)) == 0)
%if abs(f(r(n)))<=tol
break;
elseif (f(a)*f(r(n)) <= 0)
b = r(n) ;
else
a = r(n);
end
end
it_table=[r' f(r)'];
clc
disp(it_table)
figure(1),clf(1)
plot(1:numel(r),f(r))
xlabel('Iteration number'),ylabel('f(r)')

maryam
maryam on 25 Oct 2022
Find an approximation to within 0.00001 to a value in [0, 1] with 𝑓(𝑥) = 𝑒 𝑥 − 𝑥 2 + 3𝑥 − 2 = 0 Use Bisection method
  1 Comment
Walter Roberson
Walter Roberson on 25 Oct 2022
No, that will certainly not solve the question posted by Adomas

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!