min ||Ax-b||_2 with linear constraints

databricks / spark-tfocs

A Spark port of TFOCS: Templates for First-Order Conic Solvers (cvxr.com/tfocs)

Apache License 2.0

89 stars 37 forks source link

min ||Ax-b||_2 with linear constraints #29

Open ghost opened 7 years ago

ghost commented 7 years ago

Hi All, I am wondering if spark-tfocs solves a linear regression problem with linear constraints. i.e.

min ||Ax-b||_2^2 s.t. Ux<=b where U is a matrix?

I know that it would solve without those constraints, but not sure if it handles the constraints.

staple commented 7 years ago

Hi mutmurat, unfortunately TFOCS for Spark does not support quadratic programming out of the box. However if your constraints happen to be very simple ranges you may be able to implement them using projection operators. (For example see the ProjBox operator.)

ghost commented 7 years ago

staple, thank you for the answer.

akshaydse commented 8 months ago

Hi can you please help me set up contraints lasso in py spark?