out <- glm(y ~ . , family=binomial(),data=turnout) In R, above syntax works and i expected that’s similar to julia, but . is not a proper symbol in GLM.jl maybe. julia> out = glm(@formula(y ~ .), train, Binomial()) ERROR: syntax: invalid identifier name "."

Yes, still the best way. Well, the way it’s written is awkward. julia> term(:y) ~ sum(term.(names(df, Not(:y)))) FormulaTerm Response: y(unknown) Predictors: x1(unknown) x2(unknown)