New interface for summary statistics version

[gaow]

@zouyuxin although we have discussed it in person, from my experience it is always best to open new issues for new features added to keep everybody in the loop even if we do not need to discuss it with others. We then close the issue only when the feature is documented and added to unit tests. Tickets helps us keeping track of development.

Yuxin and I have currently agreed to support 3 main susie function calls:

1. susie, with input X, Y etc (already implemented).
2. susie_z, with input Z and R where R is correlation. (we might optionally take var(Y) and sample size -- still working on it)
3. susie_ss, with input betahat, sebetahat, varY, N and R where N is sample size and R can be correlation, covariance or X^TX.

The old susie_ss interface which takes X^TX, X^Y, etc, will be made an internal function.

Yuxin is working on the implementation now. We can discuss some more when she has the first draft of interface implemented and documented.

[pcarbo]

@gaow @zouyuxin Thanks for the summary! Is susie_z also working with summary statistics then? Maybe we should call the functions susie_ss_z and susie_ss_beta, or something like that?

[gaow]

Sure. Our initial thought was susie_z and susie_bs but the latter is apparently not good. @stephens999 do you have other suggestions or we use what @pcarbo proposed?

[stephens999]

maybe _bhat?

…

On Mon, Oct 8, 2018 at 8:46 AM gaow ***@***.***> wrote: Sure. Our initial thought was susie_z and susie_bs but the latter is apparently not good. @stephens999 <https://github.com/stephens999> do you have other suggestions or we use what @pcarbo <https://github.com/pcarbo> proposed? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#53 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABt4xaJBgtRUvJPCaNrwIqsra0bqpb9Aks5ui1c_gaJpZM4XMRTc> .