susie pipeline fix

[hsun3163]

No description provided.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[gaow]

thanks @hsun3163 at the face value it looks it is completed? I need to review the logic in the utils script a bit more carefully but it looks great! For this line and comment:

for (r in 1:length(fdat$residual_Y_scaled)) { ## Cant have a universal way to specify names due to the accomodation of missingness, use index instead

Can we use residual_X_scaled instead because we will always have it? I think it is important to keep the actual name of r in the RDS file that we save.

[hsun3163]

thanks @hsun3163 at the face value it looks it is completed? I need to review the logic in the utils script a bit more carefully but it looks great! For this line and comment:

for (r in 1:length(fdat$residual_Y_scaled)) { ## Cant have a universal way to specify names due to the accomodation of missingness, use index instead

Can we use residual_X_scaled instead because we will always have it? I think it is important to keep the actual name of r in the RDS file that we save.

I have tested the current pipeline last night and it runs while specifying two phenoFile and covFile, I.e two condition. I use residY b/c, when we want to run mvsusie, there will be only 1 y, and multiple residX, but we only want to run mvsusie once.

I feel like it is quite hard to specify the name of r when you need to correctly drop the name while a phenoFile is missing.

[gaow]

I feel like it is quite hard to specify the name of r when you need to correctly drop the name while a phenoFile is missing.

That is true. But that is only relevant to mvSuSiE where there will not be this line anyways: r in 1:length(fdat$residual_Y_scaled) right? So what's in SuSiE related code should work well with names