Functional Programming in R Advanced Statistical Programming for Data Science, Analysis and Finance by Unknown

Author:Unknown
Language: eng
Format: epub
Published: 0101-01-01T00:00:00+00:00

Chapter 3 ■ SCope and CloSureS

We then assign the value 3 to a local variable and get the environment chain:

["x -> 2, "result -> 4, "y -> 3] ->

["y -> 2, "f -> function(x) ...]

Now both x and y have local scope. It doesn’t matter for the result, though. We have already evaluated the expression x + y to get the result, so what we return is 4, not 6.

Situations like this don’t just happen when we assign to a local variable later in a function. If we conditionally assign to a local variable, this is also in effect. In the function below, when we evaluate x + y the scope of these parameters depends on whether we assigned to them before we evaluated the expressions. So these variables can have local or global scope depending on which parameters we called the function with: f <- function(condx, x, condy, y) {

if (condx) x <- 2

if (condy) y <- 2

x + y

}

So to briefly summarize how scopes and environments work in R: whenever you evaluate an expression, there is an associated chain of environments. The scope of the variables in the expression, which variables the actual variable names refer to, depends on where in this chain the variables can be found. While data are immutable in R, environments are not; whenever you assign to a variable with the <- operator, you are modifying the top environment in the environment chain. This can not only change the value a variable refers to but also its scope. If we assign to a variable inside a function, we are only changing the environment inside that function. Other functions that might be referring to a global variable with the same name will still be referring to the global variable, not the new local variable, because the environment that will be created when these functions are called will not be chained to the local environment where a new variable has been put.

The rules for how variables are mapped to values are always the same. It involves a search in the chain of environments that are active at the time the expression is evaluated.

The only difficulty is knowing which environments are in the chain at any given time.

Here the rules are not that complex either. We can explicitly create an environment and put it at the top of the chain of environments using the eval function, we can create a call environment when we pass expressions as arguments to a function—where the environment will be the same environment chain as where we call the function—or we can create a new environment by running code inside a function.

Scopes, Lazy Evaluation, and Default Parameters

Knowing the rules for how variables are mapped into values also helps us understand why we can use function parameters in expressions we use for default parameters when defining a function, but we cannot use them when we call a function.

51

Chapter 3 ■ SCope and CloSureS

If we define a function with a default parameter set to

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.