With Ranges, you may still want to perform computations based on the indices of the elements in the range. I wrote this minimal snippet illustrating the basic usage together with a simple example.
std::vector v{1, 2, 3, 4, 5};
for (auto&& item : v | boost::adaptors::indexed())
// item.index(), item.value()
auto range_with_preceded =
v | boost::adaptors::indexed()
| boost::adaptors::filtered(
[](auto&& item)
{
return item.index() > 0;
}
);
Look how cool the lambda is with the auto argument!! I’ve heard this being called Generic lambda.
Now I am interested in seeing if this is as efficient as I understood. I understood that ranges are “lazy”, so that for instance when I compose in the example above to create range_with_preceded, the adaptors are just “views” on the same underlying data. Maybe godbolt.org is the right place to experiment with this idea. Let me know if you have any thoughts about any of this!
Background: https://www.fluentcpp.com/2018/02/09/introduction-ranges-library/
Check out the comments section in the YouTube video for their recommendations. Boost also has Ranges.
that takes an object from the space 
and outputs and object from the space 
. In HM-SVMs the objects of 
whose definition appears in the tutorial about HM-SVM that Nico, my GSoC mentor, wrote [
. In this case, we use the Hamming distance defined for strings. In our case, the inputs of the loss function are sequences of the same length so the Hamming loss is simple the count of the number of elements that are different in the sequences (think of it as the minimum number of modifications one may do to one of the sequences to make it equal to the other. Providing that the sequences are of the same length, the only modification allowed is to change the value of an element in one of the sequences).
was zero for every problem instance. After double checking with MATLAB’s function quadprog that the correct solution for the problem was effectively non-trivial, I tried adding some lines here and there to ensure that the parameters given to MOSEK’s were the correct ones. All the matrices and vectors were ok. After some more lines to print out the bounds of the problem I found it. The variables of the optimization problem were all fixed to zero!
. However, this is false. If no bound is given, then the reality is that MOSEK fixes the value of the variable to zero. I still don’t get the point of including variables to the optimization vector whose values are fixed… but that’s another story.
for the QP with box constraints). I have started to write some code in CPrimalMosekSOSVM that makes use of MOSEK to solve the QP. This piece of code is still rather poor and it is just in my local repository.
computation and another one for the structured loss function 
.
Shogun Developer 