search

dtkaplan / statisticalModeling

Other
11 stars 17 forks source link

fmodel and effect_size() don't work for linear models with poly() terms #3

Closed dtkaplan closed 8 years ago

dtkaplan commented 8 years ago

For instance, mod <- lm(outcome == "Alive" ~ poly(age, 3) + smoker, data = Whickham) or for splines::ns().

dtkaplan commented 8 years ago

Fixed. The trick was to get the data from the data = argument in the call to lm() or glm(), just as was already being done with rpart(), etc.