{"id":436,"date":"2015-04-12T15:15:40","date_gmt":"2015-04-12T07:15:40","guid":{"rendered":"https:\/\/www.techcoil.com\/blog\/?p=436"},"modified":"2018-09-04T22:47:17","modified_gmt":"2018-09-04T14:47:17","slug":"how-to-save-and-load-environment-objects-in-r","status":"publish","type":"post","link":"https:\/\/www.techcoil.com\/blog\/how-to-save-and-load-environment-objects-in-r\/","title":{"rendered":"How to save and load environment objects in R"},"content":{"rendered":"

There was a need for me to build a prediction model in R and a Shiny app to allow users to get predictions out of my model. As the building of a prediction model take quite a while, it is not feasible for my shiny app to build the prediction model on demand. <\/p>\n

There must be a way for my Shiny app to load my prediction model only once for fulfilling prediction requests from the users. As with many other programming languages, there are mechanisms in R that allow me to save my environment objects in one session and load them back in another session. This post documents how I can save and load environment objects in R.<\/p>\n

Getting some objects into the environment<\/h2>\n

To demonstrate the ability of R in saving and loading objects, I first create some objects into the environment. <\/p>\n

\r\nx = c(1,2,3)\r\ny = c('Hello', 'World')\r\naList <- list()\r\naList['a'] <- 'a'\r\naList['b'] <- 'b'\r\n<\/pre>\n
Here I had created a Integer vector, a Character vector and a list of Character vectors. I then ran the following function to see if my objects had been saved to the current environment:<\/p>\n
\r\nls()\r\n<\/pre>\n
which gave me the following output:<\/p>\n
\r\n[1] "aList" "x"     "y"\r\n<\/pre>\n
Saving the entire list of environment objects<\/h2>\n
After I had created some objects into the R environment, I used the save.image<\/code> function to help me save the entire list of environment objects to a file. To do so, I ran the following function:<\/p>\n
\r\nsave.image(file='myEnvironment.RData')\r\n<\/pre>\n
This function ran the ls<\/code> function to get the entire list of objects in the environment and save the objects as a file named 'myEnvironment.RData' in my current working directory. <\/p>\n
To check if I had 'myEnvironment.RData', I ran the following function to get the files available in my current working directory:<\/p>\n
\r\ndir()\r\n<\/pre>\n
and was able to see the newly created file that should contain the objects that I had created earlier:<\/p>\n
\r\n[1] "myEnvironment.RData" \r\n<\/pre>\n
Loading back the entire list environment of objects<\/h2>\n
To demonstrate that I can load back the objects that I had created, I first terminate my current R session by running the following function:<\/p>\n
\r\nquit(save='no')\r\n<\/pre>\n
This function tells R to quit the current session without saving the environment data.<\/p>\n
I then fire up a new R session and run the following function to load back the objects from 'myEnvironment.RData':<\/p>\n
\r\nload('myEnvironment.RData')\r\n<\/pre>\n
This function looked up 'myEnvironment.RData' in my current working directory in order to load back the objects that are saved within the 'myEnvironment.RData' file.<\/p>\n
I ran the ls<\/code> function again to check whether my objects are loaded into the environment:<\/p>\n
\r\nls()\r\n<\/pre>\n
which produced the following output:<\/p>\n
\r\n[1] "aList" "x"     "y"\r\n<\/pre>\n\n