Multiplication of sparse 3D matrix with dense 2D decision variable matrix

Ongoing

Ker Hak

October 25, 2020 05:57 Edited

• 
• 

I have sparse COO 3D A matrix N*N*M (The third dimension is M) and dense 2D B matrix (N*N), which will be an optimization variable for my solver. I want to multiply A and B to obtain (N*N*M) matrix. Then I need to sum the elements of the resulting 3D matrix to create 2D matrix (N*N).

I am using pydata sparse library to create 3D sparse matrix.

My objective function is to minimize the sum of the elements of the multiplication of sparse A with decision variable matrix B.

How can I do this?

I used

np.sum(A*B)

 as the objective function. However, I received the following error.

TypeError: operand type(s) all returned NotImplemented
   from __array_ufunc__(<ufunc 'multiply'>, '__call__',
   <COO: shape=(5, 1454, 1454), dtype=float64, nnz=100, fill_value=0.0>,
   <(1454, 1454) matrix variable>): 'COO', 'MVar'

0

• Official comment

  

  Simranjit Kaur

  • Gurobi Staff

  June 23, 2025 03:05 Edited

  This post is more than three years old. Some information may not be up to date. For current information, please check the Gurobi Documentation or Knowledge Base. If you need more help, please create a new post in the community forum. Or why not try our AI Gurobot?.
• 

  Matthias Miltenberger

  • Gurobi Staff

  November 26, 2020 11:36 Edited

  Hi Ker,

  Please try to provide a small, reproducible example code with your data structures. PyData sparse does not seem to be a standard Python library, so your best bet is probably to loop over the matrix dimensions and create the objective function iteratively. 

  Cheers,
  Matthias

  0

Post is closed for comments.