sgd

sgd copied to clipboard

• [ ] subset: a subset of data points; can be a parameter in sgd.control
• [ ] na.action: how to deal when data has NA; can be a parameter in sgd.control
• [ ] model: logical value determining whether to output the X data frame
• [ ] x,y: logical value determining whether to output the x and/or y
• [ ] contrasts: a list for performing hypothesis testing on other sets of predictors; can be a paramter in sgd.control
• [ ] weights
• [ ] offset

I plan to add the default value to na.action, subset in valid_model_control(). However, i'll need to call valid_model_control() in sgd.formula. So valid_model_control() will be called twice. I think the redundancy is acceptable. na.action and subset will be shared by all models, so we may also add them as arguments in sgd.formula. But I think that would make sgd.formula look too complex. Do you think we need na.action and subset for sgd.matrix?

as long as there are reasonable defaults, it would make sense to include those.

It also makes sense to profile the code and see how long running the validity check takes On May 9, 2015 4:22 PM, "Panos Toulis" [email protected] wrote:

as long as there are reasonable defaults, it would make sense to include those.

— Reply to this email directly or view it on GitHub https://github.com/airoldilab/sgd/issues/48#issuecomment-100541105.

I haven't formally profiled the code. But I once ran the code with validity checks commented out. There's no significant improvement in performance.