sgd

sgd copied to clipboard

Models using boost::math::tools::schroeder_iterate can get stuck

4

Hi, I am building a package that is using sgd. When doing tests I sometimes encountered models that got stuck in the sgd C++ part for no particular reason, crashing...

jonlachmann

It seems like `model='cox'` is not actually supported because `predict.sgd` doesn't know how to handle it, despite `sgd` accepting it. https://github.com/airoldilab/sgd/blob/a256063e7add0ec43f051c2b0ba5a959e950dceb/R/predict.sgd.R#L23-L26 This then means that the example in https://github.com/airoldilab/sgd/blob/master/demo/cox-regression.R does...

deaneckles

Wine quality example

3

I am trying to SGD for classification of a huge dataset and wanted first to try it on the example you provided about the wine quality. But with the wine...

fortisil

SGD with arbitrary function for likelihood and gradient

1

Is there a way to use the sgd package for a maximum likelihood estimation that does not fit into one of the statistical families? . I see that there is...

EliYannop

No support for weights

2

Hi, it would be great to have support for a weights vector as in glmnet

jeffwong-nflx

better clarify how to do prediction in documentation

1

from an e-mail exchange with @drizzersilverberg: > 1. Can 'sgd' package perform linear regression task to build a model and predict a set of data using the built model? >...

dustinvtran

clarify how to stream data to sgd

1

I noticed in the vignette it alludes to being able to stream data to sgd via bigmemory but I can't seem to find any examples of how to do this....

alexanderchernyakovgithub

Similar to [`update.nlme`](https://www.rdocumentation.org/packages/nlme/versions/3.1-1/topics/update.nlme). Ideally it would really enable SGD updates passing in new data at a time. A first implementation of the method can just re-run SGD from iteration 1.

dustinvtran

`coef(sgd.theta)` returns the vector. In general, it seems convention is to return a vector and not a matrix. Should we change `sgd$coefficients` to do so too? This is a user-breaking...

dustinvtran

We have seem this behavior when using constant learning rates with logistic regression models. In theory the iterates should not diverge with constant rates but bounce around the true value....

ptoulis