Home

Intro to R

• Jump to ...
  • Home
  • Part 0
  • Part 1
  • Part 2
  • Part 3
  • Part 4
  • Part 5
  • Part 6
  • Part 7
  • Part 8
  • Answers to tasks

---

R Tutorial Part 7: Saving results

1 Recap

In the part 7 we learned:

• If statements to run some code only if a certain condition holds.
• For loops to run some code several times for each item in a vector.
• Lists, which are sequences of vectors, where each vector can have a different length.

2 Saving a data frame

The best way to save results is in universally readable formats, for instance csv files for data or pdf files for graphs (see the previous parts). However, R does allow saving objects directy, which is sometimes useful with very complex data types.

Data frames can be written to a csv file like this:

write.csv(d, file="../data/myNewData.csv")

If you’re dealing with linguistic data, like IPA text, you might need to state that you want to write to UTF-8, which preserves special characters in their correct format (particularly characters with accents or diachrotics, but also emojis and special math characters):

write.csv(d, file="../data/myNewData.csv", fileEncoding = "UTF-8")

3 Saving an R object

You can use the save function to save any object/variable directly. The code below will create a file called “secretCodes.Rdat” which will store a variable. No need to do this, just get a sense of how it works:

secretCodes <- c(1,6,3,8,10)
save(secretCodes, file="../data/mySuperSecrets.Rdat")

We can load the file again (maybe in a different script), and the object will be re-loaded into the memory.

load("../data/mySuperSecrets.Rdat")
secretCodes

## [1]  1  6  3  8 10

Note that the name of the file and the name of the object don’t have to match up.

4 Saving a workspace

It’s possible to save all objects loaded into the memory (the workspace) in a single file that you can re-open to pick up where you left off. RStudio will give you an option to save the workspace each time it quits.

While this can be convenient, it’s not recommended. If an object is required by the script, then it should be created by the script. It’s easy to create an object, save it in the workspace, but forget to add the lines of code to the script. When developing scripts for scientific research, you want the code to be reproducible by others. So forcing yourself to run the script from scratch a few times is good practice.

---

Go to the next tutorial

Back to the index