python - 在 cvxopt 中显示最优解

Question

有没有办法只显示cvxopt中的最佳值而不显示迭代值，例如 pcost、dcost、gap 等，如下例所示，

n = 5 #no. of design variables
P = spmatrix(1,range(n),range(n))
q = matrix([[60.341,13.05,1.217,5.5,24.2]])
m = matrix([[0,0,-2.778,-0.002657,0,0.3004,0],[0.6168,0,0,-0.0005693,0,0.06217,0],[0,5,0,0,0,0,0],[-0.8987,0,0,0,-0.000462,0,0.1554],[0,0,3.33,0,-0.001943,0,0.683]])
nl = spmatrix(-1,range(n),range(n))
nu = spmatrix(1,range(n),range(n))
G = matrix([m,nl,nu])
h = matrix([0.0113,1,0.167,0.20289,0.5129,0.9098,0.829,0.05,0.45,0,0.35,0,0.1,1.55,10,1.4,0.15])
qp = solvers.qp(P,q,G,h)
print qp['x']

输出：

    pcost       dcost       gap    pres   dres
0: -1.0569e+03 -4.5953e+02  3e+03  6e+00  2e+00
1: -1.1312e+02 -1.5294e+02  6e+02  8e-01  2e-01
2: -1.8148e-01 -5.6851e+01  6e+01  2e-16  7e-15
3: -6.5682e+00 -1.6549e+01  1e+01  9e-17  7e-16
4: -1.0162e+01 -1.1053e+01  9e-01  3e-16  2e-16
5: -1.0470e+01 -1.0624e+01  2e-01  6e-17  3e-16
6: -1.0502e+01 -1.0506e+01  4e-03  1e-16  4e-16
7: -1.0503e+01 -1.0503e+01  4e-05  7e-17  2e-16
8: -1.0503e+01 -1.0503e+01  4e-07  1e-16  2e-16
Optimal solution found.
[-5.00e-02]
[-4.50e-01]
[ 1.97e-08]
[-3.21e-01]
[ 7.81e-10]

我只想打印输出的最后 5 个值。

Answer 1

只需按照文档中的说明关闭详细模式（在 cvxopt 中调用show_progress）。

from cvxopt import spmatrix, matrix, solvers
solvers.options['show_progress'] = False        # !!!

n = 5 #no. of design variables
P = spmatrix(1,range(n),range(n))
q = matrix([[60.341,13.05,1.217,5.5,24.2]])
m = matrix([[0,0,-2.778,-0.002657,0,0.3004,0],
[0.6168,0,0,-0.0005693,0,0.06217,0],[0,5,0,0,0,0,0],
[-0.8987,0,0,0,-0.000462,0,0.1554],[0,0,3.33,0,-0.001943,0,0.683]])
nl = spmatrix(-1,range(n),range(n))
nu = spmatrix(1,range(n),range(n))
G = matrix([m,nl,nu])
h = matrix([0.0113,1,0.167,0.20289,0.5129,0.9098,0.829,0.05,0.45,0,0.35,0,0.1,1.55,10,1.4,0.15])
qp = solvers.qp(P,q,G,h)
print(qp['x'])

输出：

[-5.00e-02]
[-4.50e-01]
[ 1.97e-08]
[-3.21e-01]
[ 7.81e-10]