no output from verbose option. Scs problem or cvxr ? or else?

[reumle]

Hello, thank you for the great work!
(first GitHub post of my life :)

• Under plain R, when I call scs like this solut2 = psolve( prob1,verbose=FALSE,solver="SCS",max_iters=20000,warm_start=FALSE)

I do not get any output such as (from an Rstudio session) ….

----------------------------------------------------------------------------
	SCS v2.1.1 - Splitting Conic Solver
	(c) Brendan O'Donoghue, Stanford University, 2012
----------------------------------------------------------------------------
Lin-sys: sparse-direct, nnz in A = 6015
eps = 1.00e-005, alpha = 1.50, max_iters = 20000, normalize = 1, scale = 1.00
acceleration_lookback = 10, rho_x = 1.00e-003
Variables n = 505, constraints m = 1011
Cones:	linear vars: 1011
Setup time: 1.22e-003s
----------------------------------------------------------------------------
 Iter | pri res | dua res | rel gap | pri obj | dua obj | kap/tau | time (s)
----------------------------------------------------------------------------
     0|3.82e+020 2.01e+020 1.00e+000 -1.74e+023 1.52e+020 3.51e+023 9.52e-005 
   100|6.81e-003 1.52e-002 2.25e-003 8.29e+001 8.25e+001 9.14e-014 4.24e-002 
   200|1.08e-003 3.81e-003 6.66e-004 8.28e+001 8.27e+001 1.01e-013 8.31e-002 
….

• and especially not error metrics such as

----------------------------------------------------------------------------
Status: Solved
Timing: Solve time: 1.24e+000s
	Lin-sys: nnz in L factor: 10041, avg solve time: 1.85e-005s
	Cones: avg projection time: 5.81e-007s
	Acceleration: avg step time: 3.90e-004s
----------------------------------------------------------------------------
Error metrics:
dist(s, K) = 1.5322e-015, dist(y, K*) = 0.0000e+000, s'y/|s||y| = 2.7636e-016
primal res: |Ax + s - b|_2 / (1 + |b|_2) = 1.4158e-006
dual res:   |A'y + c|_2 / (1 + |c|_2) = 7.6361e-006
rel gap:    |c'x + b'y| / (1 + |c'x| + |b'y|) = 1.0912e-006
----------------------------------------------------------------------------
c'x = 82.6803, -b'y = 82.6805
============================================================================

• Under Rstudio the printout hangs in the middle. But I can get it (eventually) by calling the solver again (such as see below, from Rstudio console):

  2100|2.22e-005 1.94e-005 5.30e-006 8.27e+001 8.27e+001 1.33e-013 8.69e-001 
  2200|2.11e-005 1.54e-005 8.72e-007 8.27e+001 8.27e+001 3.67e-014 9.17e-001 
  2300|2.05e-005 1.29e-005 5.29e-007 8.27e+001 8.27e+001 5.05e-014 9.64e-001 
  
> solut2	=	psolve( prob1,verbose=FALSE,solver="SCS",max_iters=20000,warm_start=FALSE)	
> 
> solut2	=	psolve( prob1,verbose=FALSE,solver="SCS",max_iters=20000,warm_start=FALSE)	
> 
> solut2	=	psolve( prob1,verbose=TRUE,solver="SCS",max_iters=20000,warm_start=FALSE)	
2400|2.30e-002 1.72e-002 2.69e-004 8.26e+001 8.27e+001 2.33e-014 1.01e+000 
  2500|1.82e-005 1.50e-005 6.01e-006 8.27e+001 8.27e+001 1.65e-014 1.06e+000 
  2600|2.20e-005 1.80e-005 3.31e-007 8.27e+001 8.27e+001 7.10e-014 1.12e+000 
  2700|2.71e-005 1.34e-004 3.20e-006 8.27e+001 8.27e+001 1.36e-014 1.16e+000 
  2800|6.47e-006 1.41e-005 5.42e-007 8.27e+001 8.27e+001 5.70e-014 1.20e+000 
  2900|1.42e-006 7.64e-006 1.09e-006 8.27e+001 8.27e+001 7.07e-014 1.24e+000 
----------------------------------------------------------------------------
Status: Solved
Timing: Solve time: 1.24e+000s
	Lin-sys: nnz in L factor: 10041, avg solve time: 1.85e-005s
	Cones: avg projection time: 5.81e-007s
	Acceleration: avg step time: 3.90e-004s
----------------------------------------------------------------------------
Error metrics:
dist(s, K) = 1.5322e-015, dist(y, K*) = 0.0000e+000, s'y/|s||y| = 2.7636e-016
primal res: |Ax + s - b|_2 / (1 + |b|_2) = 1.4158e-006
dual res:   |A'y + c|_2 / (1 + |c|_2) = 7.6361e-006
rel gap:    |c'x + b'y| / (1 + |c'x| + |b'y|) = 1.0912e-006
----------------------------------------------------------------------------
c'x = 82.6803, -b'y = 82.6805
============================================================================

• Session info for the plain R session.

> sessionInfo()
R version 3.6.1 (2019-07-05)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 17763)
….
other attached packages:
 [1] scs_1.3-2         CVXR_0.99-7       tidyr_1.0.0       dplyr_0.8.3       tibble_2.1.3      R.utils_2.9.0     R.oo_1.23.0       R.methodsS3_1.7.1 stringr_1.4.0     Matrix_1.2-17    

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.3       magrittr_1.5     bit_1.1-14       tidyselect_0.2.5 lattice_0.20-38  R6_2.4.1         rlang_0.4.2      tools_3.6.1      grid_3.6.1       bit64_0.9-7      assertthat_0.2.1 lifecycle_0.1.0 
[13] crayon_1.3.4     gmp_0.5-13.5     ECOSolveR_0.5.3  purrr_0.3.3      vctrs_0.2.0      zeallot_0.1.0    glue_1.3.1       stringi_1.4.3    Rmpfr_0.7-2      compiler_3.6.1   pillar_1.4.2     backports_1.1.5 
[25] pkgconfig_2.0.3 
> 

I cannot be sure this is due to scs as opposed to CVXR.. what do you think?
Best regards, let me know if I can help.

[bnaras]

Please post with a reproducible example to the CVXR issues page.