ECOS v2.0.4 released

[staticfloat]

Hello there, I've been going through updating binary dependencies on Homebrew; is there a reason to disallow ECOS v2.0.4? Can we relax the version constraint at all?

[mlubin]

@tkelman, mind bumping the windows binaries?

[tkelman]

I'm getting what looks like a test failure on win32

=================================== log(ax)-x ===================================

ECOS 2.0.4 - (C) embotech GmbH, Zurich Switzerland, 2012-15. Web: www.embotech.com/ECOS

It     pcost       dcost      gap   pres   dres    k/t    mu     step   sigma     IR    |   BT
 0  +0.000e+00  -3.542e-01  +3e+00  6e-01  5e-01  1e+00  1e+00    ---    ---    0  0  - |  -  -
 1  -2.218e+00  -1.244e+00  +6e-01  3e-01  3e-01  2e+00  2e-01  0.7833  9e-03   1  1  1 |  1  1
 2  -1.463e+00  -9.302e-01  +1e-01  1e-01  9e-02  8e-01  4e-02  0.7833  9e-03   1  1  1 |  1  1
 3  -8.061e-01  -6.379e-01  +2e-02  3e-02  2e-02  2e-01  8e-03  0.7833  9e-03   1  1  1 |  1  1
 4  -7.249e-01  -6.910e-01  +4e-03  7e-03  5e-03  5e-02  2e-03  0.7784  1e-02   1  1  1 |  1  1
 5  -6.864e-01  -6.773e-01  +9e-04  2e-03  1e-03  1e-02  4e-04  0.7833  9e-03   1  1  1 |  1  1
 6  -6.822e-01  -6.804e-01  +2e-04  4e-04  3e-04  2e-03  1e-04  0.7733  2e-02   1  1  1 |  1  1
 7  -6.803e-01  -6.798e-01  +5e-05  9e-05  7e-05  6e-04  2e-05  0.7833  9e-03   2  1  1 |  1  1
 8  -6.800e-01  -6.799e-01  +1e-05  2e-05  2e-05  1e-04  5e-06  0.7746  2e-02   1  1  0 |  1  1
 9  -6.799e-01  -6.799e-01  +3e-06  5e-06  4e-06  3e-05  1e-06  0.7833  9e-03   1  0  0 |  1  1
10  -6.799e-01  -6.799e-01  +7e-07  1e-06  8e-07  7e-06  3e-07  0.7752  2e-02   1  0  0 |  1  1
11  -6.799e-01  -6.799e-01  +1e-07  2e-07  2e-07  2e-06  6e-08  0.7833  9e-03   1  0  0 |  1  1
12  -6.799e-01  -6.799e-01  +4e-08  6e-08  4e-08  3e-07  2e-08  0.7833  1e-02   1  0  0 |  1  1
13  -6.799e-01  -6.799e-01  +6e-09  1e-08  1e-08  8e-08  3e-09  0.7833  9e-03   2  0  0 |  1  1
14  -6.799e-01  -6.799e-01  +9e-10  1e-09  4e-10  2e-09  3e-10  0.9791  9e-03   0  0  0 |  1  0
15  -6.799e-01  -6.799e-01  +2e-10  3e-10  2e-10  2e-09  8e-11  0.4010  3e-01   1  0  0 |  5  4
16  -6.799e-01  -6.799e-01  +4e-11  7e-11  5e-11  4e-10  2e-11  0.7833  9e-03   0  0  0 |  1  1
17  -6.799e-01  -6.799e-01  +1e-11  2e-11  1e-11  9e-11  4e-12  0.7833  9e-03   0  0  0 |  1  1
18  -6.799e-01  -6.799e-01  +2e-12  4e-12  2e-12  2e-11  9e-13  0.7833  9e-03   1  0  0 |  1  1
19  -6.799e-01  -6.799e-01  +5e-13  8e-13  5e-13  4e-12  2e-13  0.7833  9e-03   1  0  0 |  1  1

OPTIMAL (within feastol=8.4e-13, reltol=7.5e-13, abstol=5.1e-13).
Runtime: 0.000326 seconds.

log_ax_x: ECOS failed to produce the desired precision
Tests run: 27

[tkelman]

The failing test was added in the most recent commit - embotech/ecos@3595371. May be worth compiling on 32 bit Linux to see if it's not just a Windows problem.

[chriscoey]

@mlubin what's the status on this? am I slash should I be using the new version of ecos for my experiments?

[mlubin]

Everything should already be bumped to 2.0.4?

[chriscoey]

ok so this issue can be closed?

[mlubin]

Well except for the windows binaries

[mlubin]

Closed in favor of #47