[mlpack-git] [mlpack] Serialization of ann (#513)

[stereomatchingkiss]

Some problems need to be solved.

Optimizer of ann will take the address of DecomposableFunctionType, the solution I come up is 

```
void Serialize(Archive& ar, const unsigned int /* version */)
  {
    using mlpack::data::CreateNVP;

    ar & CreateNVP(outSize, "outSize");
    ar & CreateNVP(bias, "bias");
    ar & CreateNVP(weights, "weights");
    ar & CreateNVP(delta, "delta");
    ar & CreateNVP(gradient, "gradient");
    ar & CreateNVP(inputParameter, "inputParameter");
    ar & CreateNVP(outputParameter, "outputParameter");
    ar & CreateNVP(optimizer, "optimizer");    
    ar & CreateNVP(ownsOptimizer, "ownsOptimizer");

    //after serialize optimizer, 
    optimizer.DecomposableFunction(*this);
  }

```

optimizer part

  ```
template<typename Archive>
  void Serialize(Archive& ar, const unsigned int /* version */)
  {
    using mlpack::data::CreateNVP;

    //do not serialize function, the address of the function
    //should be setup by the DecomposableFunction
    //ar & CreateNVP(function, "function");
    ar & CreateNVP(lr, "lr");
    ar & CreateNVP(alpha, "alpha");
    ar & CreateNVP(eps, "eps");
    ar & CreateNVP(meanSquaredGad, "meanSquaredGad");
    ar & CreateNVP(gradient, "gradient");
  }
```

What do you think about it?Thanks

---
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/issues/513#issuecomment-179082507
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160203/ed5b6f26/attachment.html>