MATLAB Answers

0

Calculating Residual Values in a Regression

Asked by sarath l on 10 Nov 2018 at 20:53
Latest activity Answered by Jeff Miller on 10 Nov 2018 at 23:38
I have a doubt in calculating Residual values for a given regression model. Consider this case (first table with x and y values). As we know that residual values are (y-y_hat), I calculated them using the same formula and the values i got are same as mentioned in the website i linked. But, when i try to find the residuals using fitlm/LinearModel.fit methods, I get different values. please see the code and help me where I'm wrong.
y=[70 65 70 95 85]';
x=[60 70 80 85 95]';
mdl=LinearModel.fit(y,x);
disp(mdl.Residuals.('Raw'))
The Output I'm getting for the above code is,
-12.7778
0.9524
7.2222
-6.4286
11.0317
which doesn't match with the residuals given in the website
4.5890
-6.8493
-8.2877
13.4932
-2.9452

  0 Comments

Sign in to comment.

1 Answer

Answer by Jeff Miller on 10 Nov 2018 at 23:38
 Accepted Answer

In MATLAB's linearmodel.fit command, x is the first parameter and y is the second. You have them reversed.

  0 Comments

Sign in to comment.