Advanced R by Hadley WickhamChapter 2: Names and ValuesAsmae Toumi@asmae_toumi2020-07-261 / 13

What's in Chapter 2:2 / 13

What's in Chapter 2:Section 2.2: distinction between names and values
2 / 13

What's in Chapter 2:

Section 2.2: distinction between names and values
Section 2.3: describes when R makes a copy

2 / 13

What's in Chapter 2:

Section 2.2: distinction between names and values
Section 2.3: describes when R makes a copy
Section 2.4: explores how much memory an object occupies

2 / 13

What's in Chapter 2:

Section 2.2: distinction between names and values
Section 2.3: describes when R makes a copy
Section 2.4: explores how much memory an object occupies
Section 2.5: describes the two important exceptions to copy-on-modify

2 / 13

What's in Chapter 2:

Section 2.2: distinction between names and values
Section 2.3: describes when R makes a copy
Section 2.4: explores how much memory an object occupies
Section 2.5: describes the two important exceptions to copy-on-modify
Section 2.6: concludes the chapter with a discussion of the garbage collector

2 / 13

What's in Chapter 2:

Section 2.2: distinction between names and values
Section 2.3: describes when R makes a copy
Section 2.4: explores how much memory an object occupies
Section 2.5: describes the two important exceptions to copy-on-modify
Section 2.6: concludes the chapter with a discussion of the garbage collector

Prerequisites

To understand how R represents objects, we'll need to install the lobstr package:

library(lobstr)
# We also install emo because it's fun!
# devtools::install_github("hadley/emo")

2 / 13

Binding basics

How would you read the following?

x <- c(1, 2, 3)

3 / 13

Binding basics

How would you read the following?

x <- c(1, 2, 3)

Create an object named ‘x’, containing the values 1, 2, and 3: 💩

3 / 13

Binding basics

How would you read the following?

x <- c(1, 2, 3)

Create an object named ‘x’, containing the values 1, 2, and 3: 💩
It’s creating an object, a vector of values (1, 2, 3) and it’s binding that object to a name, x : 😄

3 / 13

Copy-on-modify

x <- c(1, 2, 3)

## [1] 1 2 3

How can we see what's happening under the hood?

4 / 13

Copy-on-modify

x <- c(1, 2, 3)

## [1] 1 2 3

How can we see what's happening under the hood?

You can call obj_address() to see this object's identifier:

obj_addr(x)

## [1] "0x125d01a8"

4 / 13

Copy-on-modify

x <- c(1, 2, 3)

## [1] 1 2 3

How can we see what's happening under the hood?

You can call obj_address() to see this object's identifier:

obj_addr(x)

## [1] "0x125d01a8"

y <- x

4 / 13

Copy-on-modify

x <- c(1, 2, 3)

## [1] 1 2 3

How can we see what's happening under the hood?

You can call obj_address() to see this object's identifier:

obj_addr(x)

## [1] "0x125d01a8"

y <- x

obj_addr(y)

## [1] "0x125d01a8"

4 / 13

What happens to `x` when you modify `y` ?

y[[3]] <- 4

5 / 13

What happens to `x` when you modify `y` ?

y[[3]] <- 4

## [1] 1 2 3

5 / 13

What happens to `x` when you modify `y` ?

y[[3]] <- 4

## [1] 1 2 3

Changing y did not modify x.

5 / 13

What happens to `x` when you modify `y` ?

y[[3]] <- 4

## [1] 1 2 3

Changing y did not modify x.

This is due to a behavior called copy-on-modify.

5 / 13

What happens to `x` when you modify `y` ?

y[[3]] <- 4

## [1] 1 2 3

Changing y did not modify x.

This is due to a behavior called copy-on-modify.

obj_addr(x)

## [1] "0x125d01a8"

obj_addr(y)

## [1] "0x13293038"

5 / 13

What about functions?

The same copy-on-modify behavior applies for functions.

6 / 13

What about functions?

The same copy-on-modify behavior applies for functions.

We can use tracemem() to track when an object gets copied. It allows us to do that because every time an object gets copied, a message containing the address of the object will be printed.

6 / 13

What about functions?

The same copy-on-modify behavior applies for functions.

We can use tracemem() to track when an object gets copied. It allows us to do that because every time an object gets copied, a message containing the address of the object will be printed.

f <- function(a) {
  a
}

x <- c(1, 2, 3)
cat(tracemem(x), "\n")

## <000000001E0C8120>

6 / 13

What about functions?

The same copy-on-modify behavior applies for functions.

We can use tracemem() to track when an object gets copied. It allows us to do that because every time an object gets copied, a message containing the address of the object will be printed.

f <- function(a) {
  a
}

x <- c(1, 2, 3)
cat(tracemem(x), "\n")

## <000000001E0C8120>

z <- f(x)

We got no message here, which means no new copy was generated.

6 / 13

What about functions?

The same copy-on-modify behavior applies for functions.

We can use tracemem() to track when an object gets copied. It allows us to do that because every time an object gets copied, a message containing the address of the object will be printed.

f <- function(a) {
  a
}

x <- c(1, 2, 3)
cat(tracemem(x), "\n")

## <000000001E0C8120>

z <- f(x)

We got no message here, which means no new copy was generated.

If f did modify x, then a new copy would get generated and thus a message would get printed by tracemem().

6 / 13

Lists

Like vectors, lists also use copy-on-modify behaviour.

7 / 13

Lists

Like vectors, lists also use copy-on-modify behaviour.

list_1 <- list(1, 2, 3)

list_2 <- list_1

7 / 13

Lists

Like vectors, lists also use copy-on-modify behaviour.

list_1 <- list(1, 2, 3)

list_2 <- list_1

obj_addr(list_1)

## [1] "0x121f6fc0"

obj_addr(list_2)

## [1] "0x121f6fc0"

7 / 13

Lists

Like vectors, lists also use copy-on-modify behaviour.

list_1 <- list(1, 2, 3)

list_2 <- list_1

obj_addr(list_1)

## [1] "0x121f6fc0"

obj_addr(list_2)

## [1] "0x121f6fc0"

list_2[[3]] <- 4

obj_addr(list_2)

## [1] "0x1e482598"

7 / 13

Lists (continued)

We can use lobstr::ref() to print the memory address of each object along with a local ID so that we can easily cross-reference shared components.

8 / 13

Lists (continued)

We can use lobstr::ref() to print the memory address of each object along with a local ID so that we can easily cross-reference shared components.

ref(list_1, list_2)

## o [1:0x121f6fc0] <list> 
## +-[2:0x1db8e278] <dbl> 
## +-[3:0x1db8e240] <dbl> 
## \-[4:0x1db8e208] <dbl> 
##  
## o [5:0x1e482598] <list> 
## +-[2:0x1db8e278] 
## +-[3:0x1db8e240] 
## \-[6:0x1e289378] <dbl>

This shows that list_1 and list_2 have shared components, namely integers 2 and 3 corresponding to the 2nd and 3rd element in their vectors.

8 / 13

Data Frames

Data frames are lists of vectors.

9 / 13

Data Frames

Data frames are lists of vectors.

If you modify a column:
- only that column needs to be modified
- the others will still point to their original references:

If you modify a row:
- every column is modified
- every column must be copied:

9 / 13

Character Vectors

Consider this character vector:

x <- c("marco", "polo", "marco", "polo")

ref(x, character = T)

## o [1:0x13428ae0] <chr> 
## +-[2:0x137efa98] <string: "marco"> 
## +-[3:0x137ef9f0] <string: "polo"> 
## +-[2:0x137efa98] 
## \-[3:0x137ef9f0]

This is called a global string pool where each element of a character vector is a pointer to a unique string in the pool. This has implications for how much memory a character vector uses. To find out, use lobstr::obj_size()

10 / 13

Modify-in-place (1)

Modyfing an R object usually creates a copy. Exceptions:

11 / 13

Modify-in-place (1)

Modyfing an R object usually creates a copy. Exceptions:

objects with a single binding

11 / 13

Modify-in-place (1)

Modyfing an R object usually creates a copy. Exceptions:

objects with a single binding

Environments, a special type of object, are always modified in place (more on this in Chapter 7)

11 / 13

Modify-in-place (2)

Implication: create functions that “remember” their previous state (more on this in Chapter 16)

12 / 13

Unbinding / Garbage collector13 / 13

Unbinding / Garbage collectorObjects get deleted thanks to the garbage collector (GC) 
13 / 13

Unbinding / Garbage collectorObjects get deleted thanks to the garbage collector (GC) 
GC frees up memory by deleting R objects that are no longer used
13 / 13

Unbinding / Garbage collectorObjects get deleted thanks to the garbage collector (GC) 
GC frees up memory by deleting R objects that are no longer used
GC runs automatically whenever R needs more memory to create a new object. 
13 / 13

Unbinding / Garbage collectorObjects get deleted thanks to the garbage collector (GC) 
GC frees up memory by deleting R objects that are no longer used
GC runs automatically whenever R needs more memory to create a new object. 
There is no reason to call gc() yourself  unless you want to:ask R to return memory to your operating system so other programs can use it, or
to know how much memory is currently being used

13 / 13

↑, ←, Pg Up, k	Go to previous slide
↓, →, Pg Dn, Space, j	Go to next slide
Home	Go to first slide
End	Go to last slide
Number + Return	Go to specific slide
b / m / f	Toggle blackout / mirrored / fullscreen mode
c	Clone slideshow
p	Toggle presenter mode
t	Restart the presentation timer
?, h	Toggle this help

Advanced R by Hadley Wickham

Chapter 2: Names and Values

Asmae Toumi

@asmae_toumi

2020-07-26

What's in Chapter 2:

What's in Chapter 2:

What's in Chapter 2:

What's in Chapter 2:

What's in Chapter 2:

What's in Chapter 2:

What's in Chapter 2:

Binding basics

Binding basics

Binding basics

Copy-on-modify

Copy-on-modify

Copy-on-modify

Copy-on-modify

What happens to x when you modify y ?

What happens to x when you modify y ?

What happens to x when you modify y ?

What happens to x when you modify y ?

What happens to x when you modify y ?

What about functions?

What about functions?

What about functions?

What about functions?

What about functions?

Lists

Lists

Lists

Lists

Lists (continued)

Lists (continued)

Data Frames

Data Frames

Character Vectors

Modify-in-place (1)

Modify-in-place (1)

Modify-in-place (1)

Modify-in-place (2)

Unbinding / Garbage collector

Unbinding / Garbage collector

Unbinding / Garbage collector

Unbinding / Garbage collector

Unbinding / Garbage collector

What's in Chapter 2:

Help

What happens to `x` when you modify `y` ?

What happens to `x` when you modify `y` ?

What happens to `x` when you modify `y` ?

What happens to `x` when you modify `y` ?

What happens to `x` when you modify `y` ?