| 12345678910111213141516171819202122232425262728293031323334 |
- import numpy as np
- from scipy import sparse
- import scipy as sp
- import utils.codegen_utils as cu
- # Set numpy seed for reproducibility
- np.random.seed(2)
- n = 50
- m = 150
- # Generate random Matrices
- Pt = sparse.random(n, n)
- P = Pt.T.dot(Pt) + sparse.eye(n)
- P = sparse.triu(P, format='csc')
- q = sp.randn(n)
- A = sparse.random(m, n).tolil() # Lil for efficiency
- u = 3 + sp.randn(m)
- l = -3 + sp.randn(m)
- # Make random problem primal infeasible
- A[int(n/2), :] = A[int(n/2)+1, :]
- l[int(n/2)] = u[int(n/2)+1] + 10 * sp.rand()
- u[int(n/2)] = l[int(n/2)] + 0.5
- # Convert A to csc
- A = A.tocsc()
- # Generate problem solutions
- sols_data = {'status_test': 'primal_infeasible'}
- # Generate problem data
- cu.generate_problem_data(P, q, A, l, u, 'primal_infeasibility', sols_data)
|
