What is the difference between the regress function and the lscov function?

I would like to perform multiple linear regression, and it seems that there are two options for MATLAB functions, regress and lscov. What is the difference between these two functions? Does regress also use least square's minimization?

Answer 1

[Shashank Prasanna]

This is well explained in the documentation.

REGRESS and LSCOV use \ or mldivide to solve a least squares problem.

http://www.mathworks.com/help/matlab/math/systems-of-linear-equations.html

LSCOV:

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

[Nicole]

Where does it say that regress and lscov use mldivide? It's not on their respective documentation pages. Furthermore, if they both use mldivide, then why have two separate functions at all?

[Shashank Prasanna]

You can open the function to see what goes on under the hood:

>> edit regress, edit lscov

Also you can scroll down in lscov documentation to the first example to see the mention of \ or mldivide.

Depending on what you want to do you may choose either. If you want to do simple linear regression go with regress. LSCOV can also do weighted least squares solutions for positive semidefinite covariance matrices. LSCOV can also be used for multivariate regression. Implementation details are covered in the documentation and the references.