I'm working with very large data in R (point clouds of about 22 GB). My computer can't handle it, so I'm looking to use Google Colab for their GPUs. My problem is having to install packages and load in the data every time I start a new session. I'm looking online and can see Colab works with VMs, and they get wiped after the session is over, but I'm wondering if you're aware of any workaround for this? I want to have the packages I already installed saved somehow so all I need to do is just load them into the library (library(lidR) for instance), and I also want the data I uploaded saved so I don't have to upload a 22GB file every time. R studio, for instance, saves the "workspace" as a file on your local machine, so every time you start, it just loads from when you last stopped. Is there a way to achieve something like this? I have a deadline approaching so I'm looking for any means to work as fast as possible. Pls help advise. Continue reading...
