Question on cost

Alessandro Dal Grande

unread,

Sep 9, 2015, 10:48:09 PM9/9/15

to Ceres Solver

Hi, I was just looking at a non-linear optimization report and the reported cost is different from what I see when debugging:









Solver Summary (v 1.10.0-no_lapack-no_openmp)



                                     Original                  Reduced

Parameter blocks                            1                        1

Parameters                                  2                        2

Residual blocks                             1                        1

Residual                                    1                        1



Minimizer                        TRUST_REGION



Dense linear algebra library            EIGEN

Trust region strategy     LEVENBERG_MARQUARDT



                                        Given                     Used

Linear solver                        DENSE_QR                 DENSE_QR

Threads                                     1                        1

Linear solver threads                       1                        1



Cost:

Initial                          1.795629e+07

Final                            1.787265e+07

Change                           8.363425e+04



Minimizer iterations                       21

Successful steps                            7

Unsuccessful steps                         14



Time (in seconds):

Preprocessor                           0.0001



  Residual evaluation                  0.0000

  Jacobian evaluation                  0.0009

  Linear solver                        0.0003

Minimizer                              0.0014



Postprocessor                          0.0000

Total                                  0.0015



Termination:                      CONVERGENCE (Function tolerance reached. |cost_change|/cost: 8.800459e-07 <= 1.000000e-06)

When I debug, I have one residual of size 1, and the initial value I see is around 5000, while the final is around 4000. Why is the reported cost so different?

Thanks

Sameer Agarwal

unread,

Sep 9, 2015, 11:21:35 PM9/9/15

to Ceres Solver

Alessandro,
I do not quite understand your question. Can you elaborate please?

--
You received this message because you are subscribed to the Google Groups "Ceres Solver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ceres-solver...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ceres-solver/b88e7877-2f8e-4928-b4d5-dd4e8e04b338%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Alessandro Dal Grande

unread,

Sep 9, 2015, 11:29:06 PM9/9/15

to Ceres Solver

Hi Sameer,

it might be due to my lacking knowledge, but isn't the cost the sum of residuals? If I know that the (only) residual starts from ~5*10^3, how comes the report says ~17*10^6?

Thanks

Sameer Agarwal

unread,

Sep 9, 2015, 11:47:49 PM9/9/15

to Ceres Solver

No the cost is 1/2 times the sum of squared residuals.

Sameer

To view this discussion on the web visit https://groups.google.com/d/msgid/ceres-solver/b0c6fdcd-3fba-41ba-9dfa-0317de229e31%40googlegroups.com.

Alessandro Dal Grande

unread,

Sep 10, 2015, 12:01:41 AM9/10/15

to Ceres Solver

Ah totally forgot that. Thanks!

Reply all

Reply to author

Forward