Where to Store Your Functions

When you create a function to use in your analysis, you often create it and store it at the top of your script or .Rmd file as a first step. However, lots of functions at the top of your code can make your code dense and harder to read.

It is good practice to create separate R scripts that you can use to store sets of related functions. You can then call those functions using the source() function, at the top of your script in the same way that you call an R package. R will then load those functions into memory and you can use them!

Sourcing functions is good practice because it is:

1. Reusable: It allows you to reuse functions over and over using the same code (i.e. you don’t have to copy and paste the function into each new analysis script).
2. Easy to Maintain: It allows you to quickly fix a function that doesn’t work properly - only once.
3. Sharable: In the same way that a library can be used by anyone, you can share your R script containing your functions with anyone, too. This is the first step towards creating an R package!

How to Source Functions in R

To source a set of functions in R:

1. Create a new R Script (.R file) in the same working directory as your .Rmd file or R script. Give the file a descriptive name that captures the types of functions in the file.
2. Open that R Script file and add one or more functions to the file.
3. Save your file.

Next,

• Open your .Rmd file or R script.
• At the top of your file, add the source(path/tofile/here.R) function.

source("remote-sensing-functions.R")

If the .R script is in your main working directory, then it won’t have a path element before it like week_06/functionfile.R vs functionfile.R.

If it’s in a different directory, adjust the path accordingly. Once you run the code containing the source() function, all of the functions in your .R file will load into your global environment. You can now use them in your script!