So far, we have looked at how to optimize the computational performance of an R code. What if, after optimizing the code, it still runs too slowly? In this chapter, we will look at how to overcome the performance limitations caused by on-the-fly interpretation of an R code using a compiled code. Many CRAN packages use compiled code to offer optimum performance, so a simple way to take advantage of a compiled code is to use these packages. Sometimes, however, a specific task needs to be performed for which no package exists. It is useful to know how to write a compiled code for R in order to make R programs run faster.

We will first see how to compile R code before its execution, then we will explore how to integrate compiled languages like C/C++ into R so that we can run R programs at native CPU speed.

This chapter covers the following topics: