Useful R package for generating datasets

I wanted to let everyone know about the simstudy R package, in case you had not seen it before. It allows you to generate synthetic data sets by defining the relationships between variables, and errors on those variables.

I have found it very useful for generating a data set for testing mixed models - you can define the random effects and nesting structure quite easily, and then see if your models can recover this.

Hi Tom,

Thank you for sending us this information. It sounds a very useful package for simulations. The synthpop R package is an alternative for generating synthetic data but I’m not sure if it has the ability to generate nested data.