I have developed a linear regression model using lm. My main purpose is to predict a prediction interval using the function predict.
As it stands right now, the lm object is too big for my taste. As a result, I was wondering if there was any way to reduce the size of lm object but still be able to use the predict function.
Right now I have deleted effects, fitted.values, residuals, and model from the lm object while still maintaining functionality of the predict function. Now the lm object takes 88 mb of space, but this is still too high.
Is there any other way to reduce the size of the object even further while still being able to use predict?
