prediction() methods - to the extent possible - should gain an at argument so they behave similarly to margins(..., at= list(var = vals)). This will be very helpful for people who want to plot predicted values in the style of Stata's marginsplot after (non-derivative) uses of margins
prediction()methods - to the extent possible - should gain anatargument so they behave similarly tomargins(..., at= list(var = vals)). This will be very helpful for people who want to plot predicted values in the style of Stata'smarginsplotafter (non-derivative) uses ofmargins