Regression in R with grouped variables -


the dependent variable value of data frame df predicted using independent variables mean, x, y in following way:

df <- df %>%      group_by(country, sex) %>%     do({          mod = lm(value ~ mean + x + y, data = .)          <- predict(mod, .)         data.frame(., a)     }) 

data grouped country , sex. so, fitting formula can expressed as:

value(country, sex) = a0(country, sex) + a1(country, sex) mean + a2(country, sex) x + a3(country, sex) y 

however, want use formula:

value(country, sex) = a0(country, sex) + a1(country, sex) mean + a2(country) x + a3(country) y 

where a2 , a3 independent of sex. how can it?

i don't think can when grouping country , sex. group country , add interactions sex:

df <- df %>%  group_by(country) %>% do({      mod = lm(value ~ sex + mean*sex + x + y, data = .)      <- predict(mod, .)     data.frame(., a) }) 

or estimate model in 1 go adding interactions sex , country:

mod <- lm(value ~ sex*country*mean + country*x + country*y <- predict(mod) 

Comments

Popular posts from this blog

javascript - Karma not able to start PhantomJS on Windows - Error: spawn UNKNOWN -

c# - Display ASPX Popup control in RowDeleteing Event (ASPX Gridview) -

Nuget pack csproj using nuspec -