CODEWITHSUNDEEP
matlabmathvectorization
will_cheuk
will_cheuk

Reputation: 379

Any suggestions to find the approximated values without using for-loop?

I have defined a vector [0.11,1,3,4.5,7]. I also has a function f which have the values [f(0.11),f(1),f(3),f(4.5),f(7)]. At the same time, I have another vector which may not have fixed length. Its value and length are determined by user. For simplicity, let assume the vector to [4,4.5,5,6]. I would like to do as follow in programming:

Obtain approximated values f(4), f(4.5), f(5) and f(6) from the given data input using linear approximation.

I understand that I can use for-loop to solve the problem; but I wonder if for-loop is necessary. Could someone think of a method without using for-loop?

Upvotes: 0

Views: 81

Answers (1)

Mikhail_Sam
Mikhail_Sam

Reputation: 11238

We can do it this way:

  1. We can do it avoiding for-loop (and any other loops too :) ):

    x = [0.11 1 3 4.5 7]'
f_values = [1 5 3 15 0]'
xnew = [4 4.5 5 6]'
fnew = interp1q(x,f_values,xnew)

  2. But there is second interesting way: to fit your data. Matlab has strong fitting tool. For example this one. If you fit it you take function like this: myf = @(x) here+some+formula. So you can get any value of f for any x in anytime without any additional calculation.

  3. If you know real value of f just create an anonymous function:

    f = @(x) x.^2

    And now you can just put your [4,4.5,5,6] values and get result:

    xnew = [4 4.4 5 6]
result = f(xnew)

Upvotes: 5

Related Questions