Let’s say I have a data frame with columns c1,c2,…,cn and I run a regression of c1 on c2, c3, c4 and c2.*c4: sm = GLM.lm( @formula(c1 ~ c2 + c3 +c4 +c2&c4), df) If I were in Python, I could then do formula = 'c1 ~ c2 + c3 +c4 +c2&c4' import patsy y, X = patsy.dmatrices(formula, df) This would gi…

Running counterfactuals on a regression with interactions

tbeason March 8, 2021, 2:53am 2

I think you are looking for modelmatrix(sm)

1 Like

Topic		Replies	Views
Julia equivalent of the R function `model.matrix` General Usage statistics	2	1047	October 31, 2020
GLM is slow on large datasets. Using OnlineStats for regressions? MixedModels? Performance glm	25	5207	November 26, 2018
Get columns of DataFrames.jl ModelMatrix object Statistics	14	2921	March 24, 2017
PSA: breaking changes in StatsModels v0.6.0 (Terms 2.0: Son of Terms) Statistics	6	1618	July 2, 2019
How to get a GLM where formula is programmatically generated New to Julia dataframes , glm	3	1328	September 1, 2020