Help with implementation about roptim

[ypan1988]

Yes, it works fine with this change.

I just did not understand why this initialization was needed given that it is passed by reference.

Anyway, feel free to close if you do not want to make this change for convenience.

It seems to be a good point. I quickly checked the code in https://github.com/ypan1988/roptim/blob/master/src/functor.h

My guess is that line 155 should be changed
from arma::vec par(x, n), grad;
to arma::vec par(x, n), grad(n);

Will verify if it will solve this problem when I am available.

[privefl]

The problem seems to come from the fact that gr is of size 0 (instead of 2) in Gradient(); not sure why.

[privefl]

It seems I need to initialize gr = arma::zeros<arma::vec>(2);.

I think this could be done once internally at the beginning, for simplicity and convenience.

[ypan1988]

It seems I need to initialize gr = arma::zeros<arma::vec>(2);.

I think this could be done once internally at the beginning, for simplicity and convenience.

Have you managed to compile and run your program? I guess it might be the reason and I was doing exactly the same thing in my examples (say line 25-32 of file https://github.com/ypan1988/roptim/blob/master/src/roptim_examples.cpp)

[privefl]

Yes, it works fine with this change.

I just did not understand why this initialization was needed given that it is passed by reference.

Anyway, feel free to close if you do not want to make this change for convenience.