How to show exact answers in MATLAB

Hello everyone, I run into this exact answer problem with MATLAB pretty frequently. My problem is right now when I'm trying to do an inverse of a matrix containing square roots, it gives me answers in fractions or in decimals. I want it to show the square roots and other functions. How would I go about doing this? I've googled a while and all they say are precision values and rounding and what not, but nothing to do with these square roots.

[Karan Gill]

Use symbolic arithmetic. See: https://www.mathworks.com/help/symbolic/choose-symbolic-or-numeric-arithmetic.html

As Walter answered, use

>> inv([sqrt(sym(21))])
ans =
21^(1/2)/21

[Mr.Gh]

Use this code to see the exact answer:

>>A = inv([sqrt(sym(21))]);

>>answer = double(A)

answer = 2.1822e-01

Answer 1

[Mahdi]

Why don't you try an approach where you use square the matrix to find out what the numbers are?

For example

MatrixSquared=Matrix.^2;

Then you would be able to see which numbers were square rooted. I would also suggest placing

format rat

before the code.

[Suh-Shin]

Hey, I tried your approach, but it only gives me fractions. I want MATLAB to show me square roots and powers. I tried the following:

format rat;
inv([sqrt(21)]);

it shows me fractions, but I want to see square roots

Answer 2

[Matt J]

Sounds like you're trying to do a symbolic matrix inverse operation. The Symbolic Math Toolbox will let you do that, see

http://www.mathworks.com/help/symbolic/inv.html

[Suh-Shin]

Hi

I just tried the inv you've talked about, however, when I try the following, I still do not get square roots. I want the answer to be in square roots:

inv([sqrt(21)])

[Walter Roberson]

inv([sqrt(sym(21))])

[Seetha Rama Raju Sanapala]

I want to get answers in fractional form for operations on fractions like 1/2 + 15/17, 2/5*3/16 etc. Which route is better - sym or format rat? Why?

[Walter Roberson]

sym as early as possible, before you even do the division. For example sym(15)/sym(17). Otherwise you are losing precision in the fractions. format rat does not return the true fraction corresponding to the internal representation of the floating point numbers: format rat uses a continued fraction approach to try to find the closest fraction within a tolerance. So where true fraction might be 38183/86917, format rat might decide that 38/87 is "close enough" to display

[Seetha Rama Raju Sanapala]

Great clarity, @ Walter Roberson! Thanks.

[Karan Gill]

To add onto Walter's reply, this topic explains differences in arithmetic.

Answer 3

[Shashank Prasanna]

Use format to reset your display format or choose a different display format:

>> format

http://www.mathworks.com/help/matlab/ref/format.html

[Suh-Shin]

format doesn't show the answer in square roots.