NAIS/CrossEntropy: add samples accessors for each step with selection

[jschueller]

What is the idea?

Accessors could be added, see SubsetSampling methods getIn|OutputSample(k, select) for step k and select flag

Why is this needed?

No response

Additional Context

No response

[m-balesdent]

Currently, these samples can be accessed with an old fashion way : see this example

I agree a dedicated accessor could be better !

Do you mean accessors for both input and output samples ?

[jschueller]

input/output, per step and per failed/ok

[m-balesdent]

Hi @jschueller ,

I am going to change NAIS / Cross Entropy algorithms to add samples accessors.

In subsetsampling, the samples for each step are currently saved inside the subsetsampling class (and accessors are present in the subsetsampling) but I'm wondering if it could be more convenient for the user to set the acessors in the subsetsampling results class instead of subsetsampling. What is your opinion on that topic?

Thanks

[jschueller]

this is the right thing to do unfortunately it cannot be exposed as the result is currenlty only exposed via the parent class,
so adding methods wont be easy, thats probably why I added it to the algo class

[m-balesdent]

OKay, I will do the same.