what's bad about saving the workspace to .RData?

Why is it counteradvised to save the environment to the .RData file in general? What is the benefit of the .RData file?

andresrcs in this answer and thadley in this answer recommend against using .RData and it bugs me.

It can make your work not (easily) reproducible if your previous session is restored. Also if you had a large amount of data the start-up can be very long.


Also, if one object in your current environment is the cause for a session crash by restoring it on startup you would get into and endless crash cycle and, some objects are just pointers for data stored in other tools like Java libraries and by loading those objects the R package that works as an interface also gets loaded, one example of this is the h2o package.


This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.