Example: Minimizing 1-norm subject to a 2-norm constraint

[Johan Norlinder]

Hi,

I am using the example code provided from: Minimizing 1-norm subject to a 2-norm constraint. 

(primal) minimize || u ||_1

subject to || A * u - b ||_2 <= 1

With good success, very nice :)

However, I need to speed up the computation a little. The A matrix is dense and is of dimension 16x122. Are there any "easy" extra tricks left to use in this case? 
I am solving the same problem many times with different b's. Maybe there is some matrix factorization one can reuse or choose a more clever initialization?

How would one actually set an initial point in this case(in the code)? 

Lastly, is it possible to multiply A with a weight matrix W using the same code?

Best regards,
Johan

[Martin]

You could try the following: