retrieve solution when termination_status(model) = MathOptInterface.SLOW_PROGRESS

[cxqdssaw]

Hi,

I am trying to solve an SDP problem using JuMP+Mosek. The solver terminates with

ITE PFEAS    DFEAS    GFEAS    PRSTATUS   POBJ              DOBJ              MU       TIME  
0   1.1e+00  1.0e+00  1.0e+00  0.00e+00   1.000000000e+00   1.000000000e+00   1.0e+00  0.00  
1   3.8e-01  3.5e-01  2.3e-01  5.30e-01   1.045240379e+00   1.057176508e+00   3.5e-01  0.00  
2   1.6e-01  1.5e-01  1.0e-01  2.44e-01   1.030594117e+00   1.216796797e+00   1.5e-01  0.00  
3   6.7e-02  6.2e-02  3.4e-02  2.75e-01   1.009269556e+00   1.151929356e+00   6.2e-02  0.01  
4   3.6e-02  3.3e-02  2.5e-02  8.58e-02   1.015853624e+00   1.402753602e+00   3.3e-02  0.01  
5   9.8e-03  9.0e-03  6.5e-03  -1.47e-01  1.005778137e+00   1.442817551e+00   9.0e-03  0.01  
6   2.3e-03  2.1e-03  2.1e-03  -2.27e-01  1.002995182e+00   1.882748489e+00   2.1e-03  0.01  
7   6.1e-04  5.6e-04  9.4e-04  -5.62e-01  1.002530789e+00   3.793440183e+00   5.6e-04  0.01  
8   1.5e-04  1.3e-04  4.7e-04  -8.92e-01  1.002708189e+00   1.292989084e+01   1.3e-04  0.01  
9   3.8e-05  3.5e-05  2.3e-04  -9.78e-01  1.002417191e+00   4.557452984e+01   3.5e-05  0.01  
10  9.7e-06  8.9e-06  1.2e-04  -1.00e+00  1.002261444e+00   1.738859128e+02   8.9e-06  0.01  
11  2.5e-06  2.3e-06  5.9e-05  -9.96e-01  1.002176008e+00   6.733929483e+02   2.3e-06  0.01  
12  6.4e-07  5.8e-07  3.0e-05  -1.01e+00  1.002198714e+00   2.667069495e+03   5.8e-07  0.02  
13  1.6e-07  1.5e-07  1.5e-05  -9.96e-01  1.002140216e+00   1.051271637e+04   1.5e-07  0.02  
14  4.1e-08  3.8e-08  7.7e-06  -1.01e+00  1.002185416e+00   4.192762282e+04   3.8e-08  0.02  
15  1.0e-08  9.5e-09  3.9e-06  -9.94e-01  1.002111802e+00   1.673395232e+05   9.5e-09  0.02  
16  2.6e-09  2.4e-09  2.0e-06  -1.01e+00  1.002165375e+00   6.791268510e+05   2.4e-09  0.02  
17  6.4e-10  5.8e-10  9.7e-07  -9.94e-01  1.002088309e+00   2.754515902e+06   5.8e-10  0.02  
18  1.6e-10  1.4e-10  4.8e-07  -1.01e+00  1.002135697e+00   1.131472290e+07   1.4e-10  0.02  
19  3.8e-11  3.5e-11  2.4e-07  -9.94e-01  1.002068682e+00   4.611588277e+07   3.5e-11  0.02  
20  9.5e-12  8.7e-12  1.2e-07  -1.01e+00  1.002104026e+00   1.891178125e+08   8.7e-12  0.02  
21  2.3e-12  2.2e-12  5.9e-08  -9.94e-01  1.002057982e+00   7.688155577e+08   2.1e-12  0.02  
22  5.8e-13  5.5e-13  2.8e-08  -1.00e+00  1.002065203e+00   3.134996150e+09   5.3e-13  0.03  
23  1.4e-13  1.4e-13  1.4e-08  -1.01e+00  1.001932559e+00   1.287692894e+10   1.3e-13  0.03  
24  4.1e-14  4.0e-14  6.8e-09  -9.91e-01  1.002392858e+00   5.237257834e+10   3.2e-14  0.03  
25  1.3e-14  1.8e-14  6.9e-10  -1.00e+00  1.002672928e+00   2.287471046e+11   7.3e-15  0.03  
26  2.9e-15  1.6e-14  9.2e-10  -9.99e-01  1.002075172e+00   8.008735996e+11   2.1e-15  0.03  
27  2.9e-15  1.6e-14  9.2e-10  -1.00e+00  1.002075172e+00   8.008735996e+11   2.1e-15  0.03  
28  2.9e-15  1.6e-14  9.2e-10  -1.00e+00  1.002075172e+00   8.008735996e+11   2.1e-15  0.04  
Optimizer terminated. Time: 0.04    

termination_status(model) = MathOptInterface.SLOW_PROGRESS
primal_status(model) = MathOptInterface.NO_SOLUTION
raw_status(model) = "Mosek.MSK_RES_TRM_STALL"

However, I know the optimization is approaching optimality as the log indicates (the optimal value should be 1.0), can I still retrieve a feasible solution or the primal objective value in this case? I tried ``objective_value(model)'' to retrieve the primal objective value, but it is 0.0, not 1.0 (what I want).

Thank you very much!