f_able for use in multiple imputed dataset

[Rajini Nagrani]

Hello all,

I would like to test non-linear association in my multiple-imputed dataset. Does anyone know if f_able works for multiple-imputed dataset?

I tried a couple of other things, such as
1. npregress, which I guess is not suitable for multiple-imputed dataset.
2. mkspline2 followed by mfxrcspline, but unable to get the desired graph of absolute values of outcome
3. makespline (Stata 18), but I couldn't find any documentation if margins recognizes makespline

So I would now like to give a try with f_able, it worked well for me with complete case analysis, but does anyone know if it works for multiple imputed dataset?

Thank you for any advice.

Best regards,
Rajini

[Rajini Nagrani]

Update: It worked. I just followed the very useful post #5 from here:
Estimating Margins and Marginal Effects with Restricted Cubic Splines - Statalist

and replaced margins with mimrgns