Running polygenic with no covariates, passing a string as the formula

[jmarten]

Hello all,

I have a slightly infuriating problem.

I want to run polygenic to correct for relatedness on a trait with no covariates. Ordinarily, I'd run it like this:

Code: Select all

a <- polygenic(eGFR, data=data,kin=kin)

That works fine. However, I have a lot of these to run and I don't want to hard-code everything, so I've got a script that gets the trait to run as a string.

Code: Select all

traitName <- "eGFR"

Ideally I want to use the value of the traitName variable as the formula.

Code: Select all

> a <- polygenic(traitName, data=data,kin=kin)
Error in polygenic(traitName, data = df2, kin = gkin) :
  dimension of outcome and kinship.matrix do not match

as.formula doesn't help me here as that needs a right hand side

Code: Select all

> as.formula(traitName)
Error in eval(expr, envir, enclos) : object 'eGFR_overall' not found

And using ~1 as a RHS doesn't work in polygenic:

Code: Select all

form <- as.formula(paste0(traitName, "~1"))
> a <- polygenic(form,data=df2,kin=gkin)
Error in polygenic(form, data = df2, kin = gkin) :
  All covariates have only one value


I'm at my wits' end here. Does anyone know how to get around this?

Thanks