Scribd
Upload
12 views

Numerical Methodsactivity 2

1. The document describes creating a function to find the roots of a quadratic equation. It then uses the function to find roots and compares the results to expected values, finding small errors for the naive method and large errors for Scilab's method. 2. It explores why 0.1 cannot be represented exactly in binary by showing that 0.1 rounded to 53 bits is not equal to 1-0.9, and calculating the relative error. 3. It further explains that 0.1 is approximated in binary by dividing the significant bits by a power of 2, leading to a small difference from the expected 0.1 value.

Uploaded by

tirsollantada
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
12 views

Numerical Methodsactivity 2

1. The document describes creating a function to find the roots of a quadratic equation. It then uses the function to find roots and compares the results to expected values, finding small errors for the naive method and large errors for Scilab's method. 2. It explores why 0.1 cannot be represented exactly in binary by showing that 0.1 rounded to 53 bits is not equal to 1-0.9, and calculating the relative error. 3. It further explains that 0.1 is approximated in binary by dividing the significant bits by a power of 2, leading to a small difference from the expected 0.1 value.

Uploaded by

tirsollantada
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

1.

Create a function to find the roots of a quadratic equation


function r=myroots(p);
c=coeff(p,0);
b=coeff(p,1);
a=coeff(p,2);
r(1)=(-b+sqrt(b^2-4*a*c))/(2*a);
r(2)=(-b-sqrt(b^2-4*a*c))/(2*a);
endfunction

-->exec('C:\Users\L-7\Documents\act 2 1.sce', -1)


Warning : redefining function: myroots

2.

Determine the values of the roots as expected, using naive method & scilab method & find the error
a. p=poly([-0.0001 10000.0 0.0001],"x","coeff");
e1=1e-8;
roots1=myroots(p);
r1=roots1(1);
roots2=roots(p);
r2=roots2(1);
error1=abs(r1-e1)/e1;
error2=abs(r2-e1)/e1;
printf("Expected:%e\n",e1);
printf("Naive method:%e(error=%e)\n",r1,error1);
printf("Scilab method:%e(error=%e)\n",r2,error2);
-->exec('C:\Users\L-7\Documents\act2 2.sce', -1)
Expected:1.000000e-08
Naive method:9.094947e-09(error=9.050530e-02)
Scilab method:-1.000000e+08(error=1.000000e+16)
b. e=1.e-155;
a=1;
b=1/e;
c=1;
p=poly([c b a],"x","coeff");
expected=[-e;-1/e];
roots1=myroots(p);
roots2=roots(p);
error1=abs(roots1-expected)/norm(expected);
error2=abs(roots2-expected)/norm(expected);
printf("Expected:%e%e\n",expected(1),expected(2));
printf("Naive method:%e%e(error=%e%e)\n",roots1(1),roots1(2),error1(1),error1(2));
printf("Scilab method:%e%e(error=%e%e)\n",roots2(1),roots2(2),error2(1),error2(2));
-->exec('C:\Users\L-7\Documents\act 2 3.sce', -1)
Expected:-1.000000e-155-1.000000e+155
Naive method:Inf-Inf(error=InfInf)
Scilab method:-1.000000e+155-1.000000e-155(error=1.000000e+001.000000e+00)

3.

Why is 0.1 rounded


a. format(25)

x1=0.1

x1 =0.1000000000000000055511

x2=1.0-0.9

x2 = 0.0999999999999999777955

error=(x1-x2)

error = 0.0000000000000000277556

relativeerror=(x1-x2)/x1

relativeerror = 0.0000000000000002775558

x1==x2

ans = F

b. format(25)
x=0.1

x =0.1000000000000000055511

p=53

p =53.

e=floor(log2(x))

e =- 4.

M=x/2^(e-p+1)

M =7205759403792794.

x1=720579403792793*2^(-4-53+1)

x1 = 0.0100000480645178330130

x2=720579403792794*2^(-4-53+1)

x2 =0.0100000480645178468908

0.1-x1

ans = 0.0899999519354821725381

0.1-x2

ans =0.0899999519354821586603

You might also like