aslam_optimizer

aslam_optimizer copied to clipboard

@huberya , this works on my side. Let's see what Jenkins says...

HannesSommer

see #52 .

HannesSommer

see PR #53.

HannesSommer

Remove / deprecate the evaluate*(jc, applyCR) functions

1

They are outdated thanks to the new matrix stack (PR #137 ). See 97b43ebf471f928cd87cfb4822503bc1d042308d as an example of how remove the awful duplication that always come with those as a...

HannesSommer

How hard would it be to add column normalization to the sparse cholesky linear solver?

3

furgalep

This may not make sense for the incremental calibration project but there is nothing logically wrong with it. Currently we are implementing a small class for trust region methods and...

furgalep

Feature/stochastic gradient descent

4

uschwes

Make the optimizer warn if no design variables are active

1

simonlynen

Jerome introduced already a assertion to at least informative fail the other case with d1d42388c1cae3f9b75fefce5139b0f22f4c8172.

HannesSommer

When using an M-Estimator, the wrong cost is used

5

We are not using \rho(x) as the cost. We are using w(x) x^T x. This should be changed.

furgalep