RMSE between two variables

Hello, I am really new at matlab. I am trying to create a sub-function that has an input of two vectors and output the RMSE between the values in the vectors. Anyone can help? I would love to understand step by step. thanks to anybody that can help!!

Answer 1

[Star Strider]

UPDATE: Starting in R2022b, you can now calculate Root Mean Square Error using the built in MATLAB function ‘rmse’:

https://www.mathworks.com/help/matlab/ref/rmse.html

*************************************************************

‘RMSE’ of course means ‘root mean squared error’, or the square root of the mean of the squared error.

The simplest code for this is then:

V1 = rand(10,1);
V2 = rand(10,1);

RMSE = sqrt(mean((V1-V2).^2));

where the error is (V1-V2), and ‘.^2’ denotes element-wise squaring of the error (the difference between ‘V1’ and ‘V2’). The rest of the expression takes the mean of the squared differences, and sqrt takes the square root, completing the definition.

See: Array vs. Matrix Operations (link) and Vectorization (link) for more information.

[silvia battistella]

thank you so much!!!

[Star Strider]

As always, my pleasure!

[Abinaya G EMBDD]

what is v1 and v2

[Star Strider]

Anything you want them to be!

[MAT-Magic]

@ Dear Star Strider can we use the below formula for V1 and V2?

difference = single(1) - single(2);

squaredError = difference .^ 2;

meanSquaredError = sum(squaredError(:)) / numel(signal(1));

rmsError = sqrt(meanSquaredError);

Thanks

[Star Strider]

That appears to be the correct way to calculate RMSE.

[MAT-Magic]

Thanks. I edited the question and removed the mistakes.

[Star Strider]

My pleasure.

I assume ‘single(1)’ and ‘single(2)’ are actually equal-length vectors.

[Image Analyst]

There is an rms() function that you can use. It's been in the Signal Processing Toolbox since R2012a.

[MAT-Magic]

@Star Yes, it's right.

@Image Thanks for the reply.

[Rachel Hall]

@Image Analyst,

I don't think RMSE and RMS are the same; I'm currently fining that out...

https://en.wikipedia.org/wiki/Root_mean_square

https://en.wikipedia.org/wiki/Root-mean-square_deviation

[Marcos Conde]

Yes and No. Actually, RMSE=rms(E), since E(rror) is the difference. In the example above: RMSE=rms(V1-V2)

Answer 2

[Rik]

To learn Matlab, you can use a guide like this one.

There are two main ways of doing this: an anonymous function and a 'normal' function.

%anonymous function:
calculate_RMSE=@(a,b) sqrt(mean((a(:)-b(:)).^2));

%normal function (save this in calc_RMSE.m)
function rmse=calc_RMSE(a,b)
rmse=sqrt(mean((a(:)-b(:)).^2));

The two function can be used in the exact same way. The second option provides more options for checking if the input is correct.

[silvia battistella]

thank you so much!! it really helped me!