Dtjytyk

Why do we say ‘pairwise disjoint’, rather than ‘disjoint’?

Called into a meeting and told we are being made redundant (laid off) and "not to share outside". Can I tell my partner?

What ability score modifier does a javelin's damage use?

Specifying a starting column with colortbl package and xcolor

Create chunks from an array

Why does liquid water form when we exhale on a mirror?

Why is there an extra space when I type "ls" in the Desktop directory?

Drawing close together horizontal lines in Latex

What problems would a superhuman have who's skin is constantly hot?

Having the player face themselves after the mid-game

How do spaceships determine each other's mass in space?

Minimizing with differential evolution

Does "Until when" sound natural for native speakers?

How to resolve: Reviewer #1 says remove section X vs. Reviewer #2 says expand section X

Is a piano played in the same way as a harmonium?

Virginia employer terminated employee and wants signing bonus returned

Does an unused member variable take up memory?

how to modify custom status text color in UI component grid magento 2?

How to write a chaotic neutral protagonist and prevent my readers from thinking they are evil?

How to check whether module is loaded with custom configurations?

Signed and unsigned numbers

Conservation of Mass and Energy

Was it really inappropriate to write a pull request for the company I interviewed with?

What's the 'present simple' form of the word "нашла́" in 3rd person singular female?

Why would sequential quadratic programming fail to find global minimum?

Issues with quadratic programmingQuadratic programming with simplex constraintsIs this a quadratic programming problem?SVM and quadratic programmingQuadratic programmingProve SVM Quadratic Programming has Hessian positive semidefiniteConstrained quadratic programming with positive semidefinite matrixWhat is a Sequential Quadratic Programming?Support Vector Machines: question about the underlying mathHow Sequential Quadratic Programming versus Quadratic programming and Iterative QP are related?

0

$begingroup$

I have a data set. A matrix $X$, $1300 times 20$ and output vector $mathbf{y} in Bbb R^{20}$ $$mathbf{y} = begin{bmatrix} 100\100\vdots\100end{bmatrix}$$
I am trying to run OLS on this data with $boldsymbol{beta} geq mathbf{0}$ and additional inequality constraints.

I first attempted to solve without the inequality constraints with a Python function that uses coordinate descent. This converged and this test proved I could find a global minimum. But the code doesn't accept constraints beyond $boldsymbol{beta} geq mathbf{0}$, so it's not enough for my purposes.

I then tried another Python function that uses sequential quadratic programming (SQP), which didn't converge to the global minimum. It should produce the same result since it's the same problem, but it didn't. You can see the details of my code here:

how to stop fmin_slqsp from converging to local minimum?

I determined it's likely the SQP method reached a local minimum whereas the coordinate descent method didn't.

Why would sequential quadratic programming fail to find global minimum?

positive-semidefinite quadratic-programming

share|cite|improve this question

edited 2 days ago

Stan Shunpike

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438

$endgroup$

• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  Why don't you use CVXPY?
  $endgroup$
  – Rodrigo de Azevedo
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  @RodrigodeAzevedo That looks promising. Do you have any recommendations for which methods specifically I should look at that would be useful for my case?
  $endgroup$
  – Stan Shunpike
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  That is the whole point of CVXPY. You don't worry about methods. You just write up your optimization problem and CVXPY takes it from there. If the problem is not convex, CVXPY will complain. Good for prototyping.
  $endgroup$
  – Rodrigo de Azevedo
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  $begingroup$
  It works much better but the solution is still off by a bit. It's 1444.482627745685.
  $endgroup$
  – Stan Shunpike
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  $begingroup$
  @RodrigodeAzevedo Yes, this worked. This solved the issue. Fantastic, thank you so much for your help.
  $endgroup$
  – Stan Shunpike
  2 days ago
  
  
  

  
  
  
  

 | 
show 7 more comments

0

$begingroup$

I have a data set. A matrix $X$, $1300 times 20$ and output vector $mathbf{y} in Bbb R^{20}$ $$mathbf{y} = begin{bmatrix} 100\100\vdots\100end{bmatrix}$$
I am trying to run OLS on this data with $boldsymbol{beta} geq mathbf{0}$ and additional inequality constraints.

I first attempted to solve without the inequality constraints with a Python function that uses coordinate descent. This converged and this test proved I could find a global minimum. But the code doesn't accept constraints beyond $boldsymbol{beta} geq mathbf{0}$, so it's not enough for my purposes.

I then tried another Python function that uses sequential quadratic programming (SQP), which didn't converge to the global minimum. It should produce the same result since it's the same problem, but it didn't. You can see the details of my code here:

how to stop fmin_slqsp from converging to local minimum?

I determined it's likely the SQP method reached a local minimum whereas the coordinate descent method didn't.

Why would sequential quadratic programming fail to find global minimum?

positive-semidefinite quadratic-programming

share|cite|improve this question

edited 2 days ago

Stan Shunpike

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438

$endgroup$

• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  Why don't you use CVXPY?
  $endgroup$
  – Rodrigo de Azevedo
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  @RodrigodeAzevedo That looks promising. Do you have any recommendations for which methods specifically I should look at that would be useful for my case?
  $endgroup$
  – Stan Shunpike
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  That is the whole point of CVXPY. You don't worry about methods. You just write up your optimization problem and CVXPY takes it from there. If the problem is not convex, CVXPY will complain. Good for prototyping.
  $endgroup$
  – Rodrigo de Azevedo
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  $begingroup$
  It works much better but the solution is still off by a bit. It's 1444.482627745685.
  $endgroup$
  – Stan Shunpike
  2 days ago
  

  
  
  
  


• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  $begingroup$
  @RodrigodeAzevedo Yes, this worked. This solved the issue. Fantastic, thank you so much for your help.
  $endgroup$
  – Stan Shunpike
  2 days ago
  
  
  

  
  
  
  

 | 
show 7 more comments

$begingroup$

I have a data set. A matrix $X$, $1300 times 20$ and output vector $mathbf{y} in Bbb R^{20}$ $$mathbf{y} = begin{bmatrix} 100\100\vdots\100end{bmatrix}$$
I am trying to run OLS on this data with $boldsymbol{beta} geq mathbf{0}$ and additional inequality constraints.

I first attempted to solve without the inequality constraints with a Python function that uses coordinate descent. This converged and this test proved I could find a global minimum. But the code doesn't accept constraints beyond $boldsymbol{beta} geq mathbf{0}$, so it's not enough for my purposes.

I then tried another Python function that uses sequential quadratic programming (SQP), which didn't converge to the global minimum. It should produce the same result since it's the same problem, but it didn't. You can see the details of my code here:

how to stop fmin_slqsp from converging to local minimum?

I determined it's likely the SQP method reached a local minimum whereas the coordinate descent method didn't.

Why would sequential quadratic programming fail to find global minimum?

positive-semidefinite quadratic-programming

share|cite|improve this question

edited 2 days ago

Stan Shunpike

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438

$endgroup$

share|cite|improve this question

edited 2 days ago

Stan Shunpike

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438

share|cite|improve this question

edited 2 days ago

Stan Shunpike

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438

asked 2 days ago

Stan ShunpikeStan Shunpike

1,81611438