[mlpack-git] [mlpack/mlpack] DatasetMapper & Imputer (#694)

[Ryan Curtin]

The reason I put it in the `Impute()` function itself is because it is only useful for the case where we are using only one matrix (i.e. input matrix = output matrix).  When we are using two matrices, we need to scan the whole matrix anyway, so `dirtyDimensions` won't help us there.  I think it's a little odd to offer an API that allows a user to call `Impute()` with an input and output matrix, or to call `Impute()` with a matrix and a list of dirty dimensions.  I think the list of dirty dimensions shouldn't be required as an input parameter, since the imputer already has a `DatasetMapper` object it can use to calculate the dirty dimensions.  Hopefully what I've written makes sense... let me know if I can clarify.

---
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/694#issuecomment-234614033
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160722/fe254146/attachment.html>