Intro to iteration with Base RDaniel AndersonWeek 2, Class 11 / 58

Agenda

For loops
Apply family of loops
- lapply()
- sapply()
- vapply()

2 / 58

Agenda

For loops
Apply family of loops
- lapply()
- sapply()
- vapply()

Note - we won't get to apply or tapply, but the former in particular is probably worth investigating.

2 / 58

Learning objectives

Understand the basics of what it means to loop through a vector
Begin to recognize use cases
Be able to apply basic for loops and write their equivalents with lapply.

3 / 58

Basic overview: `for` loops

4 / 58

Basic overview: `for` loops

a <- letters[1:26]
a

##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s" "t" "u"
## [22] "v" "w" "x" "y" "z"

4 / 58

Basic overview: `for` loops

a <- letters[1:26]
a

##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s" "t" "u"
## [22] "v" "w" "x" "y" "z"

for(i in 1:5){
    print(a[i])
}

## [1] "a"
## [1] "b"
## [1] "c"
## [1] "d"
## [1] "e"

4 / 58

Basic overview: `for` loops

a <- letters[1:26]
a

##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s" "t" "u"
## [22] "v" "w" "x" "y" "z"

for(i in 1:5){
    print(a[i])
}

## [1] "a"
## [1] "b"
## [1] "c"
## [1] "d"
## [1] "e"

Note these are five different character scalars (atomic vectors of length one). It is NOT a single vector.

4 / 58

Another basic example

Simulate tossing a coin, record results

5 / 58

Another basic example

Simulate tossing a coin, record results

For a single toss

sample(c("Heads", "Tails"), 1)

## [1] "Tails"

5 / 58

Another basic example

Simulate tossing a coin, record results

For a single toss

sample(c("Heads", "Tails"), 1)

## [1] "Tails"

For multiple tosses, first allocate a vector with length equal to the number of iterations

result <- rep(NA, 10)
result

##  [1] NA NA NA NA NA NA NA NA NA NA

5 / 58

Next, run the trial $n$ times, storing the result in your pre-allocated vector.

for(i in seq_along(result)) {
    result[i] <- sample(c("Heads", "Tails"), 1)
}
result

##  [1] "Tails" "Heads" "Tails" "Tails" "Heads" "Heads" "Heads" "Heads" "Heads" "Tails"

6 / 58

Growing vectors

Always pre-allocate a vector for storage before running a for loop.

7 / 58

Growing vectors

Always pre-allocate a vector for storage before running a for loop.
Contrary to some opinions you may see out there, for loops are not actually slower than lapply, etc., provided the for loop is written well

7 / 58

Growing vectors

Always pre-allocate a vector for storage before running a for loop.
Contrary to some opinions you may see out there, for loops are not actually slower than lapply, etc., provided the for loop is written well
This primarily means not growing a vector

7 / 58

Example

100,000 coin flips by growing a vector

library(tictoc)
set.seed(1)
tic()
not_allocated <- sample(c("Heads", "Tails"), 1)
for(i in seq_len(1e5 - 1)) {
    not_allocated <- c(
      not_allocated, 
      sample(c("Heads", "Tails"), 1)
     )
}
toc()

## 37.572 sec elapsed

8 / 58

same exact thing with pre-allocated vector

set.seed(1)
tic()
allocated <- rep(NA, 1e5)
for(i in seq_len(1e5)) {
    allocated[i] <- sample(c("Heads", "Tails"), 1)
}
toc()

## 0.698 sec elapsed

9 / 58

Result

The result is the same, regardless of the approach (notice I forced the random number generator to start at the same place in both samples)

identical(not_allocated, allocated)

## [1] TRUE

Speed is obviously not identical

10 / 58

You try

Base R comes with letters and LETTERS

Make an alphabet of upper/lower case. For example, create "Aa" with paste0(LETTERS[1], letters[1])
Write a for loop for all letters

03:00

11 / 58

Answer

alphabet <- rep(NA, length(letters))
for(i in seq_along(alphabet)) {
    alphabet[i] <- paste0(LETTERS[i], letters[i])
}
alphabet

##  [1] "Aa" "Bb" "Cc" "Dd" "Ee" "Ff" "Gg" "Hh" "Ii" "Jj" "Kk" "Ll" "Mm" "Nn" "Oo" "Pp" "Qq"
## [18] "Rr" "Ss" "Tt" "Uu" "Vv" "Ww" "Xx" "Yy" "Zz"

12 / 58

Quick style note

Why am I always using seq_along?

13 / 58

Quick style note

Why am I always using seq_along?
When writing functions, it's safer to use seq_* because you can't always be guaranteed of the input

x <- data.frame()
1:length(x)

## [1] 1 0

seq_along(x)

## integer(0)

13 / 58

Running the loop

for(i in 1:length(x)) {
    print(letters[i])
}

## [1] "a"
## character(0)

for(i in seq_along(x)) {
    print(letters[i])
}

The first may return unhelpful error messages or unexpected output, while the latter simply won't run, which is generally easier to diagnose.

14 / 58

Running the loop

for(i in 1:length(x)) {
    print(letters[i])
}

## [1] "a"
## character(0)

for(i in seq_along(x)) {
    print(letters[i])
}

The first may return unhelpful error messages or unexpected output, while the latter simply won't run, which is generally easier to diagnose.
Even better, if you're using a loop in a function, you should probably have a condition that checks the input before running it

14 / 58

Another example

Say we wanted to simulate 100 cases from random normal data, where we varied the standard deviation in increments of 0.2, ranging from 1 to 5

15 / 58

Another example

Say we wanted to simulate 100 cases from random normal data, where we varied the standard deviation in increments of 0.2, ranging from 1 to 5
First, specify a vector standard deviations

increments <- seq(1, 5, by = 0.2)

15 / 58

Another example

Say we wanted to simulate 100 cases from random normal data, where we varied the standard deviation in increments of 0.2, ranging from 1 to 5
First, specify a vector standard deviations

increments <- seq(1, 5, by = 0.2)

Next, allocate a vector. There are many ways I could store this result (data frame, matrix, list). I'll do it in a list.

simulated <- vector("list", length(increments))
str(simulated)

## List of 21
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL
##  $ : NULL

15 / 58

Write `for` loop

for(i in seq_along(simulated)) {
    simulated[[i]] <- rnorm(100, 0, increments[i]) 
    # note use of `[[` above
}
str(simulated)

## List of 21
##  $ : num [1:100] -2.387 0.405 -1.599 -0.285 0.288 ...
##  $ : num [1:100] 0.298 0.433 -1.021 1.384 -0.323 ...
##  $ : num [1:100] 0.893 -1.799 -0.819 -1.11 -2.198 ...
##  $ : num [1:100] -0.332 1.067 -0.823 2.899 1.863 ...
##  $ : num [1:100] -2.568 -0.672 -0.244 -1.645 2.221 ...
##  $ : num [1:100] 2.4 -1.95 1.13 3.05 3.56 ...
##  $ : num [1:100] -2.978 0.798 2.212 2.15 -2.197 ...
##  $ : num [1:100] -0.211 -1.768 3.35 2.06 0.213 ...
##  $ : num [1:100] 0.718 -4.029 -1.093 0.417 -3.952 ...
##  $ : num [1:100] 0.632 3.084 -2.62 -1.282 -2.965 ...
##  $ : num [1:100] 2.1759 -0.4681 1.6349 -0.0809 -0.7611 ...
##  $ : num [1:100] 1.236 3.055 -2.575 -0.868 4.369 ...
##  $ : num [1:100] 0.7795 -1.0125 -6.465 0.0926 1.8629 ...
##  $ : num [1:100] 3.466 -1.245 0.496 3.67 -2.207 ...
##  $ : num [1:100] -2.712 -4.21 -3.686 -0.728 -0.142 ...
##  $ : num [1:100] 2.83 6.08 -3 4.29 4.18 ...
##  $ : num [1:100] 0.335 0.574 4.106 4.414 0.897 ...
##  $ : num [1:100] -2.123 3.165 1.104 -4.065 0.578 ...
##  $ : num [1:100] 2.448 -1.472 4.411 2.34 -0.346 ...
##  $ : num [1:100] 0.672 -4.724 3.378 -1.811 10.33 ...
##  $ : num [1:100] -3.46 2.83 -11.49 -1.86 7.54 ...

16 / 58

List/data frame

Remember, if all the vectors of our list are the same length, it can be transformed into a data frame.
First, let's provide meaningful names

names(simulated) <- paste0("sd_", increments)
sim_d <- data.frame(simulated)
head(sim_d)

##         sd_1     sd_1.2     sd_1.4     sd_1.6     sd_1.8       sd_2     sd_2.2
## 1 -2.3872613  0.2979273  0.8930471 -0.3319310 -2.5676229  2.3954045 -2.9775420
## 2  0.4051212  0.4329239 -1.7989656  1.0673114 -0.6722755 -1.9542566  0.7980123
## 3 -1.5992856 -1.0209222 -0.8192303 -0.8232251 -0.2435614  1.1309855  2.2118028
## 4 -0.2847246  1.3838266 -1.1097486  2.8991565 -1.6445420  3.0545722  2.1497809
## 5  0.2881735 -0.3233308 -2.1982580  1.8633398  2.2213515  3.5578607 -2.1969056
## 6  0.1175257 -0.4150724  0.6353818  0.7142314 -0.5219616 -0.8029945  0.1464460
##       sd_2.4     sd_2.6     sd_2.8        sd_3     sd_3.2      sd_3.4     sd_3.6
## 1 -0.2105980  0.7175135  0.6321448  2.17589282  1.2362364  0.77947966  3.4658329
## 2 -1.7675364 -4.0290962  3.0843027 -0.46812673  3.0554725 -1.01245886 -1.2451870
## 3  3.3501949 -1.0927053 -2.6196216  1.63492841 -2.5751022 -6.46499466  0.4960868
## 4  2.0601974  0.4174713 -1.2824915 -0.08085208 -0.8678742  0.09259855  3.6701383
## 5  0.2125117 -3.9521276 -2.9646399 -0.76111234  4.3687915  1.86290325 -2.2067855
## 6  1.7822910 -0.1081454  4.5420524  3.53122922 -2.4194781 -1.14660593 -1.7557261
##       sd_3.8      sd_4     sd_4.2    sd_4.4     sd_4.6     sd_4.8       sd_5
## 1 -2.7119876  2.834307  0.3347596 -2.122876  2.4479578  0.6723229  -3.464762
## 2 -4.2104097  6.083823  0.5735697  3.164909 -1.4721622 -4.7240885   2.825094
## 3 -3.6861183 -3.000288  4.1063037  1.104068  4.4111688  3.3779921 -11.486532
## 4 -0.7277885  4.293730  4.4141833 -4.064636  2.3400425 -1.8110032  -1.857547
## 5 -0.1416683  4.178344  0.8965279  0.578414 -0.3461786 10.3301558   7.544994
## 6 -2.2110984  4.644032 -5.7370545 -5.072939  4.7260451  0.4489877   2.956989

17 / 58

tidyverse

One of the best things about the tidyverse is that it often does the looping for you

library(tidyverse)
pd <- sim_d %>%
    pivot_longer(
      everything(),
      names_to = "sd", 
      values_to = "sim",
      names_prefix = "sd_",
      names_ptypes = list(
        sd = factor()
      )
    ) 
ggplot(pd, aes(sim)) +
 geom_density(
   aes(color = sd)
 ) +
 guides(color = "none")

18 / 58

Of course, someone has to write loops. It doesn't have to be you. --- Jenny Bryan

19 / 58

Base R Method

Calculate all the densities

densities <- vector("list", length(sim_d))
for(i in seq_along(densities)) {
    densities[[i]] <- density(sim_d[ ,i])
}
str(densities)

## List of 21
##  $ :List of 7
##   ..$ x        : num [1:512] -3.45 -3.44 -3.42 -3.41 -3.4 ...
##   ..$ y        : num [1:512] 0.000173 0.000195 0.000219 0.000245 0.000274 ...
##   ..$ bw       : num 0.355
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -3.39 -3.38 -3.36 -3.35 -3.33 ...
##   ..$ y        : num [1:512] 0.000261 0.000296 0.000334 0.000377 0.000425 ...
##   ..$ bw       : num 0.405
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -4.9 -4.88 -4.87 -4.85 -4.83 ...
##   ..$ y        : num [1:512] 9.72e-05 1.09e-04 1.22e-04 1.37e-04 1.53e-04 ...
##   ..$ bw       : num 0.495
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -5.64 -5.62 -5.59 -5.57 -5.54 ...
##   ..$ y        : num [1:512] 0.000132 0.00015 0.000171 0.000195 0.000221 ...
##   ..$ bw       : num 0.572
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -5.97 -5.95 -5.92 -5.9 -5.88 ...
##   ..$ y        : num [1:512] 0.00012 0.000137 0.000155 0.000175 0.000198 ...
##   ..$ bw       : num 0.559
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -5.57 -5.55 -5.53 -5.5 -5.48 ...
##   ..$ y        : num [1:512] 0.000137 0.000153 0.000169 0.000187 0.000207 ...
##   ..$ bw       : num 0.698
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -7.7 -7.67 -7.64 -7.61 -7.58 ...
##   ..$ y        : num [1:512] 5.93e-05 6.65e-05 7.46e-05 8.37e-05 9.35e-05 ...
##   ..$ bw       : num 0.768
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -9.48 -9.44 -9.4 -9.37 -9.33 ...
##   ..$ y        : num [1:512] 5.38e-05 6.18e-05 7.07e-05 8.06e-05 9.17e-05 ...
##   ..$ bw       : num 0.834
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -9.05 -9.02 -8.98 -8.95 -8.91 ...
##   ..$ y        : num [1:512] 5.98e-05 6.80e-05 7.71e-05 8.74e-05 9.91e-05 ...
##   ..$ bw       : num 0.813
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -9.05 -9.02 -8.99 -8.95 -8.92 ...
##   ..$ y        : num [1:512] 6.33e-05 7.07e-05 7.87e-05 8.79e-05 9.79e-05 ...
##   ..$ bw       : num 0.954
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -9.69 -9.65 -9.61 -9.57 -9.54 ...
##   ..$ y        : num [1:512] 0.000107 0.00012 0.000135 0.000151 0.000169 ...
##   ..$ bw       : num 1.08
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -9.73 -9.69 -9.64 -9.6 -9.56 ...
##   ..$ y        : num [1:512] 9.15e-05 1.03e-04 1.15e-04 1.29e-04 1.44e-04 ...
##   ..$ bw       : num 1.12
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -11.3 -11.2 -11.2 -11.2 -11.1 ...
##   ..$ y        : num [1:512] 4.55e-05 5.14e-05 5.79e-05 6.53e-05 7.34e-05 ...
##   ..$ bw       : num 0.989
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -12 -11.9 -11.9 -11.8 -11.8 ...
##   ..$ y        : num [1:512] 4.34e-05 4.86e-05 5.45e-05 6.09e-05 6.78e-05 ...
##   ..$ bw       : num 1.04
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -11 -10.9 -10.9 -10.9 -10.8 ...
##   ..$ y        : num [1:512] 7.10e-05 7.95e-05 8.87e-05 9.88e-05 1.10e-04 ...
##   ..$ bw       : num 1.18
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -11.6 -11.5 -11.5 -11.4 -11.4 ...
##   ..$ y        : num [1:512] 6.28e-05 7.04e-05 7.89e-05 8.84e-05 9.87e-05 ...
##   ..$ bw       : num 1.3
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -15 -15 -14.9 -14.9 -14.8 ...
##   ..$ y        : num [1:512] 3.18e-05 3.64e-05 4.14e-05 4.70e-05 5.32e-05 ...
##   ..$ bw       : num 1.41
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -13.9 -13.9 -13.8 -13.8 -13.7 ...
##   ..$ y        : num [1:512] 6.35e-05 7.18e-05 8.15e-05 9.22e-05 1.04e-04 ...
##   ..$ bw       : num 1.4
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -15.6 -15.5 -15.5 -15.4 -15.4 ...
##   ..$ y        : num [1:512] 3.11e-05 3.48e-05 3.88e-05 4.33e-05 4.83e-05 ...
##   ..$ bw       : num 1.48
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -14 -13.9 -13.9 -13.8 -13.7 ...
##   ..$ y        : num [1:512] 4.97e-05 5.57e-05 6.26e-05 7.01e-05 7.83e-05 ...
##   ..$ bw       : num 1.78
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"
##  $ :List of 7
##   ..$ x        : num [1:512] -17.4 -17.3 -17.2 -17.2 -17.1 ...
##   ..$ y        : num [1:512] 3.78e-05 4.25e-05 4.76e-05 5.34e-05 5.98e-05 ...
##   ..$ bw       : num 1.71
##   ..$ n        : int 100
##   ..$ call     : language density.default(x = sim_d[, i])
##   ..$ data.name: chr "sim_d[, i]"
##   ..$ has.na   : logi FALSE
##   ..- attr(*, "class")= chr "density"

20 / 58

Next, plot the first density

plot(densities[[1]])

21 / 58

Finally, loop through all the other densities

plot(densities[[1]], xlim = c(-20, 20))
for(i in seq(2, length(densities))) {
    lines(x = densities[[i]]$x, 
          y = densities[[i]]$y)    
}

22 / 58

Skipping iterations

On the prior slide, I set the index to skip over the first by using seq(2, length(densities))

23 / 58

Skipping iterations

On the prior slide, I set the index to skip over the first by using seq(2, length(densities))
Alternatively, the loop could have been written like this

plot(densities[[1]], xlim = c(-20, 20))
for(i in seq_along(densities)) {
    if(i == 1) next
    lines(x = densities[[i]]$x, 
          y = densities[[i]]$y)    
}

23 / 58

Breaking loops

Similarly, if a condition is met, you may want to break out of the loop

set.seed(1)
rand_unif <- vector("double", 10)
for(i in seq_along(rand_unif)) {
    rand_unif[i] <- runif(1, 0, 10)
    if(any(rand_unif > 5)) {
        break
    }
}
rand_unif

##  [1] 2.655087 3.721239 5.728534 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
## [10] 0.000000

24 / 58

*apply

25 / 58

`lapply`

One of numerous functionals in R
A functional "takes a function as an input and returns a vector as output" (adv-r, Chpt 9)

26 / 58

`lapply`

One of numerous functionals in R
A functional "takes a function as an input and returns a vector as output" (adv-r, Chpt 9)
lapply will always return a list

26 / 58

Revisiting our simulation with $n = 10$

Our for loop version

increments <- seq(1, 5, by = 0.2)
simulated <- vector("list", length(increments))
for(i in seq_along(simulated)) {
    simulated[[i]] <- rnorm(10, 0, increments[i]) 
    # note use of `[[`
}
simulated

## [[1]]
##  [1]  1.329799263  1.272429321  0.414641434 -1.539950042 -0.928567035 -0.294720447
##  [7] -0.005767173  2.404653389  0.763593461 -0.799009249
## 
## [[2]]
##  [1] -1.3771884 -0.3473539 -0.3590581 -0.4938130  0.3026681 -1.0703054  0.5228200
##  [8] -1.4850461 -0.2691215  0.4528748
## 
## [[3]]
##  [1]  0.18667091  1.12586531 -0.07994948  0.70505116  1.52007711 -0.96733538 -1.79843910
##  [8]  0.06541664 -0.32998918 -0.76004356
## 
## [[4]]
##  [1] -0.6932965 -1.0391546  1.1628012  1.8430588  1.5874566 -0.6872210  1.9812866
##  [8] -0.4469541  2.8126449  0.8971937
## 
## [[5]]
##  [1] -0.8150112 -1.4976779 -2.0998270 -1.9180630 -2.8148077  2.0817666  1.4976848
##  [8] -0.4091916  0.4790473 -0.6780649
## 
## [[6]]
##  [1]  4.88272926 -1.59067823 -0.10975495  0.50028265  1.23648659 -0.34524701 -4.44780055
##  [8] -2.52722877  0.71745779 -0.02209096
## 
## [[7]]
##  [1] -2.06942816 -0.25481571 -1.79293116  0.53297966 -3.13521647  0.80507047  0.54650783
##  [8]  0.14363400  0.04214406  0.56614443
## 
## [[8]]
##  [1] -1.5576242 -0.2860050  1.5939257  2.6423258  0.3450516 -0.2826086 -2.1889641
##  [8] -3.4502070 -1.9130149  3.0097995
## 
## [[9]]
##  [1]  2.0075697 -0.5707406 -1.1045067 -1.0893483  2.5921658 -0.7170229  3.2656489
##  [8]  1.6813534  3.3782120 -2.2704815
## 
## [[10]]
##  [1]  0.02343869 -2.46644083  1.66952525  0.33520940 -0.79008686  4.07676752  0.64125485
##  [8]  2.79032300  2.18920572 -2.17497454
## 
## [[11]]
##  [1] -1.8479697  0.1397409 -3.3911573  1.7301563 -3.8422483  4.8763419 -1.5020898
##  [8]  5.0348916 -1.2375597 -2.9168605
## 
## [[12]]
##  [1]  0.08122518  0.08792107 -5.37658471  3.37200276 -3.58271713  1.07397507  1.58334645
##  [8]  0.44176867 -0.38013448  0.63258964
## 
## [[13]]
##  [1] -3.63355522 -2.73092494 -3.78680146  5.37231172  5.09258379  0.89299456 -4.19186408
##  [8] -0.01266001  5.13968576 -1.61737417
## 
## [[14]]
##  [1]  2.8724992 -3.5064092  2.4817417 -3.4410208 -4.4341454 -3.4448108 -3.1312183
##  [8] -3.2784505  2.6685947  0.2466415
## 
## [[15]]
##  [1] -1.230253 -4.128712 -3.860530 -2.917603 -4.254936 -1.703062  1.792598 -4.485865
##  [9]  5.586977 -4.983398
## 
## [[16]]
##  [1] -0.38609969  9.47887963  3.56250591 -1.00873265 -3.46305502  2.33034400 -0.05011739
##  [8] -1.49941905  1.27154294 -1.95522254
## 
## [[17]]
##  [1] 11.166364  7.057168  3.274253  2.995610 -2.280104  3.720269 -1.464098 -4.233829
##  [9]  7.909367 -3.901679
## 
## [[18]]
##  [1] -1.2944644 -2.7057812 -4.1671335  2.6354907 -6.7039055 -0.9072316 -2.5268998
##  [8] -6.1167306 -0.3098365 -1.8958699
## 
## [[19]]
##  [1] -2.72423672  4.51313434  2.44908304 -0.41609817  0.71985626 -3.39163378 -0.92616955
##  [8]  5.07001234 -0.07704198  0.74422771
## 
## [[20]]
##  [1]  9.71885467 -3.37773242  4.61180344  8.59432826 -5.10799278  0.08465542 -1.87156142
##  [8] -2.35599721 -5.01944473 -4.30181407
## 
## [[21]]
##  [1]  6.3469358  2.9692047  3.8781716  7.7868519 -1.8270090  4.0827822 -0.3031739
##  [8] -2.5068916  4.6303136  0.1846885

27 / 58

The `lapply` version

increments <- seq(1, 5, by = 0.2)
sim_l <- lapply(increments, function(sd) rnorm(10, 0, sd))
sim_l

## [[1]]
##  [1] -1.06620017 -0.23845635  1.49522344  1.17215855 -1.45770721  0.09505623  0.84766496
##  [8] -1.62436453  1.40856336 -0.54176036
## 
## [[2]]
##  [1]  0.33439767 -0.23276729  1.89138982 -1.77065716 -0.17352985 -1.14384377  0.48785128
##  [8]  2.67511464 -1.81739641 -0.07404891
## 
## [[3]]
##  [1] -0.2061791  2.1582303 -1.3745979  0.6952094  2.3757270 -0.3650308 -0.9883000
##  [8] -0.2256499  0.7018506 -1.4189555
## 
## [[4]]
##  [1]  2.583603577  0.009027176 -4.647838497 -1.771463710  2.476107092 -1.562928561
##  [7] -0.162405516  0.068240400 -2.554748823  0.785547796
## 
## [[5]]
##  [1]  0.7588861  3.3730270  1.8621258  0.1472586 -0.1485428  1.0909322 -1.5973563
##  [8]  0.1897585  0.6351741  0.9907080
## 
## [[6]]
##  [1] -2.2686619  2.9247031  1.4042334  5.0142223 -3.7800543 -1.1796256 -3.4290046
##  [8] -0.8419958  0.6202828  3.4051412
## 
## [[7]]
##  [1] -0.9754466 -2.6369136 -0.6762380  1.3663192  0.4001848  2.9004820 -0.6576005
##  [8] -3.6260878  2.0932966 -2.4488705
## 
## [[8]]
##  [1]  1.4807195  1.2323849  0.8867018  4.1373459 -0.4947470 -3.1540683  0.1523378
##  [8] -0.5567459  1.5241448  3.9231463
## 
## [[9]]
##  [1] -4.7016517 -0.5561035  0.1829520  1.4292399 -1.8117412  1.0154715  0.9916693
##  [8] -0.0321692 -0.3235310  3.8135359
## 
## [[10]]
##  [1]  1.8870003  5.4779907 -0.7533148 -3.4847443 -1.1079682  0.2727106 -0.6674835
##  [8] -1.1531183 -4.4162105 -2.2323731
## 
## [[11]]
##  [1] -3.2887103  0.9249262  1.0343854  4.6189443 -0.9885426  2.8451681 -1.4377668
##  [8] -4.5446604  1.3036100 -1.5586100
## 
## [[12]]
##  [1] -2.6705889 -2.4212723  3.4864112  5.0317853  3.2235660 -0.8741056 -4.1909745
##  [8]  0.7122363  3.5629862  2.6796621
## 
## [[13]]
##  [1]  1.0693570  0.7555414 -2.8682923  1.5089381  0.1897102  0.2311061 -0.6866699
##  [8] -3.9372519 -2.0153271  2.6046223
## 
## [[14]]
##  [1]  0.14014122  0.05272852 -0.67074108  5.04212698  0.06654805  0.89710564  0.53719465
##  [8] -3.46763945 -0.23928039  4.63291935
## 
## [[15]]
##  [1]  1.7408760 -5.5177075  0.2939007  2.1276020 -0.2847970  2.9742513 -0.6561405
##  [8] -3.9949164  2.7719149  0.9981279
## 
## [[16]]
##  [1]  2.1746314  4.1642412  0.7900246 -6.5183131  0.4841609 -6.5496878 -2.1241724
##  [8]  3.8147192 -6.8826026  0.4252825
## 
## [[17]]
##  [1] -2.5564077 -1.2650273  4.1000506  1.9152364  5.4365139 -4.7594493 -3.6517335
##  [8] -3.1708752 -0.5444685 -4.2075656
## 
## [[18]]
##  [1] -3.6074204 -4.2880309  2.6589614  2.4146654  4.0323037 11.7108920 -0.7931311
##  [8]  3.0140650 14.3722239  2.4666420
## 
## [[19]]
##  [1] -0.3174796 -4.4732375 -2.5142983 -7.7679847 -7.2329144 -1.8629409  1.4687175
##  [8]  0.1859673 -1.7940440 -8.3684223
## 
## [[20]]
##  [1]  3.1640674  2.2061840  7.7598064 -8.9097144 -1.3767546  8.4015451  0.5587853
##  [8]  6.6444152  2.7562604  0.6551559
## 
## [[21]]
##  [1]  4.57107994 -9.00413159 -1.69940320  3.03132286  6.70565155  3.83643644  0.96862833
##  [8]  5.70283345  0.06932402 -5.52652955

28 / 58

Some more examples

Loop through a data frame

Remember - a data frame is a list. We can loop through it easily

29 / 58

Some more examples

Loop through a data frame

Remember - a data frame is a list. We can loop through it easily

library(palmerpenguins)
lapply(penguins, is.double)

## $species
## [1] FALSE
## 
## $island
## [1] FALSE
## 
## $bill_length_mm
## [1] TRUE
## 
## $bill_depth_mm
## [1] TRUE
## 
## $flipper_length_mm
## [1] FALSE
## 
## $body_mass_g
## [1] FALSE
## 
## $sex
## [1] FALSE
## 
## $year
## [1] FALSE

29 / 58

lapply(mtcars, mean)

## $mpg
## [1] 20.09062
## 
## $cyl
## [1] 6.1875
## 
## $disp
## [1] 230.7219
## 
## $hp
## [1] 146.6875
## 
## $drat
## [1] 3.596563
## 
## $wt
## [1] 3.21725
## 
## $qsec
## [1] 17.84875
## 
## $vs
## [1] 0.4375
## 
## $am
## [1] 0.40625
## 
## $gear
## [1] 3.6875
## 
## $carb
## [1] 2.8125

30 / 58

Add a condition

lapply(penguins, function(x) {
    if(is.numeric(x)) {
        mean(x, na.rm = TRUE)
    }
})

## $species
## NULL
## 
## $island
## NULL
## 
## $bill_length_mm
## [1] 43.92193
## 
## $bill_depth_mm
## [1] 17.15117
## 
## $flipper_length_mm
## [1] 200.9152
## 
## $body_mass_g
## [1] 4201.754
## 
## $sex
## NULL
## 
## $year
## [1] 2008.029

31 / 58

Add a second condition

lapply(penguins, function(x) {
    if(is.numeric(x)) {
        return(mean(x, na.rm = TRUE))
    }
    else if(is.character(x) | 
                is.factor(x)) {
    return(table(x))
  }
})

## $species
## x
##    Adelie Chinstrap    Gentoo 
##       152        68       124 
## 
## $island
## x
##    Biscoe     Dream Torgersen 
##       168       124        52 
## 
## $bill_length_mm
## [1] 43.92193
## 
## $bill_depth_mm
## [1] 17.15117
## 
## $flipper_length_mm
## [1] 200.9152
## 
## $body_mass_g
## [1] 4201.754
## 
## $sex
## x
## female   male 
##    165    168 
## 
## $year
## [1] 2008.029

32 / 58

Passing arguments

head(airquality)

##   Ozone Solar.R Wind Temp Month Day
## 1    41     190  7.4   67     5   1
## 2    36     118  8.0   72     5   2
## 3    12     149 12.6   74     5   3
## 4    18     313 11.5   62     5   4
## 5    NA      NA 14.3   56     5   5
## 6    28      NA 14.9   66     5   6

lapply(airquality, mean, na.rm = TRUE)

## $Ozone
## [1] 42.12931
## 
## $Solar.R
## [1] 185.9315
## 
## $Wind
## [1] 9.957516
## 
## $Temp
## [1] 77.88235
## 
## $Month
## [1] 6.993464
## 
## $Day
## [1] 15.80392

33 / 58

Simulation again

lapply(seq(1, 5, 0.2), rnorm, n = 10, mean = 0)

## [[1]]
##  [1] -0.02516264 -0.16367334  0.37005975 -0.38082454  0.65295237  2.06134181 -1.79664494
##  [8]  0.58407712 -0.72275312 -0.62916466
## 
## [[2]]
##  [1] -2.1794473 -0.3111469  0.4015587 -1.7126011  2.3263539 -0.9114363 -2.7345314
##  [8] -0.1368609  2.8222280  1.9155850
## 
## [[3]]
##  [1]  1.7884237  1.1045592  0.6460515 -0.6132968 -2.1109298 -3.1121246 -1.6501414
##  [8] -2.4958643 -1.3830868  1.0198842
## 
## [[4]]
##  [1] -1.4154959 -2.4615063 -1.6710007 -2.7490179  1.2860121 -2.4028595 -0.2327985
##  [8]  0.9271338  1.9224409  3.0302573
## 
## [[5]]
##  [1] -3.1684074  1.6641842 -1.0017759 -0.3250514  2.6053925 -1.0928366  1.2228524
##  [8] -0.1684038 -0.8821553  2.5391869
## 
## [[6]]
##  [1] -0.4491476 -0.4249910  1.3927569  1.8303650 -1.8467486  2.2937465 -1.2717301
##  [8] -1.7728866 -4.6662734 -0.2909816
## 
## [[7]]
##  [1]  0.6973209 -1.5564340  2.7327422  1.3643348  0.2197867  3.9769477 -3.3053460
##  [8]  0.6284090  1.8605553 -2.1897538
## 
## [[8]]
##  [1] -0.6164499 -0.1340545 -1.0680126  0.1672682 -0.3712121 -1.9950323  1.8277064
##  [8] -1.3836166 -1.5032837  1.1552048
## 
## [[9]]
##  [1]  4.40770482 -4.57918836  0.51483384  1.03310766  0.07598629  6.65671081  3.26853205
##  [8] -1.38979798 -1.62559131  2.37600659
## 
## [[10]]
##  [1]  2.82015870  2.01401711 -1.69319265  1.50935234 -0.21512648  5.17977477 -2.39374114
##  [8]  0.09138443 -2.87016655 -2.75029741
## 
## [[11]]
##  [1]  0.01230587 -0.70028153 -1.49666466  4.64913889  0.26249075  3.95610339 -2.94367236
##  [8] -0.73686776 -4.21180151  4.32267944
## 
## [[12]]
##  [1] -3.14035197  4.71758369 -3.17183118 -0.30239150 -9.20045339 -0.78997152  0.04718235
##  [8] -6.14108065 -0.92100398 -1.10923983
## 
## [[13]]
##  [1] -6.2546012  3.0552024 -4.1237070 -0.7444784  1.9185117 -1.7864769  2.5308724
##  [8]  0.4385380  5.0601325 -2.2531186
## 
## [[14]]
##  [1] -4.1783580  1.2915872 -0.7014470 -1.0630152  1.7879055  1.7456860  0.0676242
##  [8]  2.2851884  2.7159987  3.0009205
## 
## [[15]]
##  [1]  3.6698929  4.9167439 -0.5188939 -1.6725270 -4.6636789 -0.9030817 -3.5220610
##  [8]  1.5626944 -0.7556854 -2.1183086
## 
## [[16]]
##  [1] -3.908628  0.242941 -2.388918 -5.035795 -5.640189  4.405216 -2.708968 -3.048694
##  [9] -1.166994 -2.300754
## 
## [[17]]
##  [1] -1.8645558 -1.3127959 -2.5326186 -4.5945258  3.0017660 -0.4570115 -6.0639513
##  [8]  3.3857177 -7.3073073 -1.6855453
## 
## [[18]]
##  [1] -1.265363 -4.128993  1.265735 -6.623765  6.684907  1.616601  7.479395  2.834467
##  [9] -7.426325  2.849642
## 
## [[19]]
##  [1]  2.064453  4.720990  4.944900  2.108224  2.905299 -2.670145  7.287284 -8.114368
##  [9] -8.650861 -5.941908
## 
## [[20]]
##  [1]   4.36641814  -5.31722727  -1.84379460   0.39712720  -2.32263588 -10.00675740
##  [7]   5.60441743  -0.36876370   2.54602270   0.02356222
## 
## [[21]]
##  [1] -2.6511571  7.5743333  4.0772982 -7.5326093 -5.7859009  6.5077747 -4.5282039
##  [8]  0.0778174 -7.1400260  3.4398985

34 / 58

Mimic `dplyr::group_by`

by_cyl <- split(mtcars, mtcars$cyl)
str(by_cyl)

## List of 3
##  $ 4:'data.frame':    11 obs. of  11 variables:
##   ..$ mpg : num [1:11] 22.8 24.4 22.8 32.4 30.4 33.9 21.5 27.3 26 30.4 ...
##   ..$ cyl : num [1:11] 4 4 4 4 4 4 4 4 4 4 ...
##   ..$ disp: num [1:11] 108 146.7 140.8 78.7 75.7 ...
##   ..$ hp  : num [1:11] 93 62 95 66 52 65 97 66 91 113 ...
##   ..$ drat: num [1:11] 3.85 3.69 3.92 4.08 4.93 4.22 3.7 4.08 4.43 3.77 ...
##   ..$ wt  : num [1:11] 2.32 3.19 3.15 2.2 1.61 ...
##   ..$ qsec: num [1:11] 18.6 20 22.9 19.5 18.5 ...
##   ..$ vs  : num [1:11] 1 1 1 1 1 1 1 1 0 1 ...
##   ..$ am  : num [1:11] 1 0 0 1 1 1 0 1 1 1 ...
##   ..$ gear: num [1:11] 4 4 4 4 4 4 3 4 5 5 ...
##   ..$ carb: num [1:11] 1 2 2 1 2 1 1 1 2 2 ...
##  $ 6:'data.frame':    7 obs. of  11 variables:
##   ..$ mpg : num [1:7] 21 21 21.4 18.1 19.2 17.8 19.7
##   ..$ cyl : num [1:7] 6 6 6 6 6 6 6
##   ..$ disp: num [1:7] 160 160 258 225 168 ...
##   ..$ hp  : num [1:7] 110 110 110 105 123 123 175
##   ..$ drat: num [1:7] 3.9 3.9 3.08 2.76 3.92 3.92 3.62
##   ..$ wt  : num [1:7] 2.62 2.88 3.21 3.46 3.44 ...
##   ..$ qsec: num [1:7] 16.5 17 19.4 20.2 18.3 ...
##   ..$ vs  : num [1:7] 0 0 1 1 1 1 0
##   ..$ am  : num [1:7] 1 1 0 0 0 0 1
##   ..$ gear: num [1:7] 4 4 3 3 4 4 5
##   ..$ carb: num [1:7] 4 4 1 1 4 4 6
##  $ 8:'data.frame':    14 obs. of  11 variables:
##   ..$ mpg : num [1:14] 18.7 14.3 16.4 17.3 15.2 10.4 10.4 14.7 15.5 15.2 ...
##   ..$ cyl : num [1:14] 8 8 8 8 8 8 8 8 8 8 ...
##   ..$ disp: num [1:14] 360 360 276 276 276 ...
##   ..$ hp  : num [1:14] 175 245 180 180 180 205 215 230 150 150 ...
##   ..$ drat: num [1:14] 3.15 3.21 3.07 3.07 3.07 2.93 3 3.23 2.76 3.15 ...
##   ..$ wt  : num [1:14] 3.44 3.57 4.07 3.73 3.78 ...
##   ..$ qsec: num [1:14] 17 15.8 17.4 17.6 18 ...
##   ..$ vs  : num [1:14] 0 0 0 0 0 0 0 0 0 0 ...
##   ..$ am  : num [1:14] 0 0 0 0 0 0 0 0 0 0 ...
##   ..$ gear: num [1:14] 3 3 3 3 3 3 3 3 3 3 ...
##   ..$ carb: num [1:14] 2 4 3 3 3 4 4 4 2 2 ...

35 / 58

lapply(by_cyl, function(x) mean(x$mpg))

## $`4`
## [1] 26.66364
## 
## $`6`
## [1] 19.74286
## 
## $`8`
## [1] 15.1

36 / 58

Your turn

Try splitting the penguins dataset by species and calculating the average bill_length_mm

04:00

37 / 58

Produce separate plots

lapply(by_cyl, function(x) {
    ggplot(x, aes(disp, mpg)) +
        geom_point() +
        geom_smooth()
})

38 / 58

## $`4`

## 
## $`6`

## 
## $`8`

## # A tibble: 344 x 8
##    species island    bill_length_mm bill_depth_mm flipper_length_mm body_mass_g sex   
##    <fct>   <fct>              <dbl>         <dbl>             <int>       <int> <fct> 
##  1 Adelie  Torgersen         39.1          18.7                 181        3750 male  
##  2 Adelie  Torgersen         39.5          17.400               186        3800 female
##  3 Adelie  Torgersen         40.300        18                   195        3250 female
##  4 Adelie  Torgersen         NA            NA                    NA          NA <NA>  
##  5 Adelie  Torgersen         36.7          19.3                 193        3450 female
##  6 Adelie  Torgersen         39.300        20.6                 190        3650 male  
##  7 Adelie  Torgersen         38.9          17.8                 181        3625 female
##  8 Adelie  Torgersen         39.2          19.6                 195        4675 male  
##  9 Adelie  Torgersen         34.1          18.1                 193        3475 <NA>  
## 10 Adelie  Torgersen         42            20.2                 190        4250 <NA>  
## # … with 334 more rows, and 1 more variable: year <int>

39 / 58

Your turn

Produce separate plots of the relation between bill_length_mm and body_mass_g

04:00

40 / 58

Saving

You can extend this example further by saving the plot outputs to an object, then looping through that object to save the plots to disk.

41 / 58

Saving

You can extend this example further by saving the plot outputs to an object, then looping through that object to save the plots to disk.
Using functionals, this would require parallel iterations, which we'll cover later (need to loop through plots and a file name)

41 / 58

Saving

You can extend this example further by saving the plot outputs to an object, then looping through that object to save the plots to disk.
Using functionals, this would require parallel iterations, which we'll cover later (need to loop through plots and a file name)
Could extend it fairly easily with a for loop

41 / 58

Saving w/`for` loop

Save plots to an object (list)

plots <- lapply(by_cyl, function(x) {
    ggplot(x, aes(disp, mpg)) +
        geom_point() +
        geom_smooth()
})

42 / 58

Saving w/`for` loop

Save plots to an object (list)

plots <- lapply(by_cyl, function(x) {
    ggplot(x, aes(disp, mpg)) +
        geom_point() +
        geom_smooth()
})

Specify file names/directory

#dir.create(here::here("plots")) 
filenames <- here::here("plots", 
                        paste0("cyl", names(by_cyl), ".png"))
filenames

## [1] "/Users/daniel/Teaching/data_sci_specialization/2020-21/c3-fp-2021/plots/cyl4.png"
## [2] "/Users/daniel/Teaching/data_sci_specialization/2020-21/c3-fp-2021/plots/cyl6.png"
## [3] "/Users/daniel/Teaching/data_sci_specialization/2020-21/c3-fp-2021/plots/cyl8.png"

42 / 58

Saving

for(i in seq_along(plots)) {
    ggsave(filenames[i], # single bracket
           plots[[i]], # double bracket
           device = "png",
           width = 6.5, 
           height = 8)
}

43 / 58

You try!

04:00

44 / 58

Variants of `lapply`

sapply
- Will try to simplify the output, if possible. Otherwise it will return a list.
- Fine for interactive work, but I strongly recommend against it if writing a function (difficult to predict the output)

45 / 58

Variants of `lapply`

sapply
- Will try to simplify the output, if possible. Otherwise it will return a list.
- Fine for interactive work, but I strongly recommend against it if writing a function (difficult to predict the output)
vapply
- Strict - you specify the output
- Use if writing functions (or just always stick with lapply), or consider jumping to {purrr} (next week)

45 / 58

Examples

Our simulation

sim_s <- sapply(seq(1, 5, by = 0.2), function(x) {
  rnorm(10, 0, x)
})
class(sim_s)

## [1] "matrix" "array"

dim(sim_s)

## [1] 10 21

sim_s

##               [,1]        [,2]       [,3]       [,4]        [,5]       [,6]       [,7]
##  [1,] -2.939773695 -0.38696441 -2.0067288 -0.6857975  0.26119194 -2.6882602 -5.3073906
##  [2,]  0.002415809 -1.60656089 -1.4772590 -1.0498872 -4.39616038 -3.0463115 -0.6903443
##  [3,]  0.509665571  0.82578723 -1.0263566  1.5350309  1.04457363 -0.8439364  3.6517332
##  [4,] -1.084720001  0.08553678  0.2952702  2.4896842  1.17909360  2.7218489  0.2880968
##  [5,]  0.704832977  2.62770283 -1.3984890 -1.6652743 -0.54811591  3.5075897  2.4109551
##  [6,]  0.330976350 -1.38924912  1.5089905  1.4889159 -1.27362282  3.1367295  1.0765501
##  [7,]  0.976327473  1.41802568 -1.6785641 -0.1207135  3.54882963  2.5935111 -1.7136026
##  [8,] -0.843339880 -0.63284203  0.3032918 -3.1475126 -0.16199763 -0.4751925  3.8358306
##  [9,] -0.970579905 -1.74795361  0.2003218 -1.2094458 -0.02523105 -2.4483003 -0.1724520
## [10,] -1.771531349  0.68756084 -1.4920501  0.7378387 -2.02222249 -0.6556254 -2.1462183
##             [,8]       [,9]     [,10]       [,11]      [,12]      [,13]      [,14]
##  [1,]  0.1695836 -1.1734150 -3.768222 -6.10501062  5.4168761 -3.3350679 -4.8069314
##  [2,] -3.6446389  2.4052205  2.894263  6.06404098  2.0548247  3.6963109  1.4332445
##  [3,]  2.0730697 -0.5164141 -2.272974  3.01792046  4.1031457  0.4737119 -0.4017125
##  [4,]  1.2037641  3.1066126  5.044831  2.45137080  0.4497503 -1.3133251  2.4326781
##  [5,] -0.8514752  1.2884162  4.960317 -1.99196485 -3.5600086  3.8201904 -2.8389526
##  [6,] -1.1722293 -5.8373967 -4.073136 -0.03384369 -1.0869663 -2.5834752 -0.3131508
##  [7,]  2.2471055 -3.4719655 -2.367832  1.85903177 -5.3272468  3.9064610  4.9762224
##  [8,] -2.5497801  3.3352155 -3.501343 -3.84371623  2.9723256 -2.8644193  0.6065646
##  [9,] -2.3611701  1.7960694  1.868407 -0.37278398  4.5338459  1.3308054  2.9634874
## [10,]  1.0181949 -2.5143629 -3.614155  0.52722496 -0.2007065  3.0306826 -0.7952206
##             [,15]      [,16]      [,17]       [,18]     [,19]       [,20]     [,21]
##  [1,] -3.91168829 -0.7117919  8.3144599  0.73152216  5.361125   0.7562907 -4.721494
##  [2,] -0.04151763 -1.7039254  3.3465371 -5.58623589 10.287085   4.4802659 -3.829450
##  [3,] -4.65496639  3.9866351 -7.1789600 10.33777061  1.390419   1.4535757 -4.768896
##  [4,] -9.86522328  2.9106428 -6.9874086 -6.21282379 -4.795530  -9.3895211 -1.990022
##  [5,]  4.44266585 -6.9065224  2.0626601 -0.07463057 -4.524295   1.6969762 -1.556085
##  [6,] -4.13025350  1.4135940 -0.7310330 -2.39500515  9.226305   2.1620377  3.980464
##  [7,] -6.93911545  2.9072547  4.0374204  7.92049427 -9.524629   3.1658442  4.932142
##  [8,]  3.78207087  2.6730439  1.2340720  4.45033678 14.056415  -4.9508195 -3.972658
##  [9,] -0.04507477 -9.6972692  0.3401973 -2.48035285 -1.202213 -11.3809098 -1.544090
## [10,] -2.27858790 -0.9414297  0.7713797  0.90385150 -2.090209  -1.5579663  1.807224

46 / 58

sapply(penguins, is.double)

##           species            island    bill_length_mm     bill_depth_mm 
##             FALSE             FALSE              TRUE              TRUE 
## flipper_length_mm       body_mass_g               sex              year 
##             FALSE             FALSE             FALSE             FALSE

47 / 58

sapply(penguins, is.double)

##           species            island    bill_length_mm     bill_depth_mm 
##             FALSE             FALSE              TRUE              TRUE 
## flipper_length_mm       body_mass_g               sex              year 
##             FALSE             FALSE             FALSE             FALSE

Now that it's a vector we can easily use it for subsetting

47 / 58

head(penguins)

## # A tibble: 6 x 8
##   species island    bill_length_mm bill_depth_mm flipper_length_mm body_mass_g sex   
##   <fct>   <fct>              <dbl>         <dbl>             <int>       <int> <fct> 
## 1 Adelie  Torgersen         39.1          18.7                 181        3750 male  
## 2 Adelie  Torgersen         39.5          17.400               186        3800 female
## 3 Adelie  Torgersen         40.300        18                   195        3250 female
## 4 Adelie  Torgersen         NA            NA                    NA          NA <NA>  
## 5 Adelie  Torgersen         36.7          19.3                 193        3450 female
## 6 Adelie  Torgersen         39.300        20.6                 190        3650 male  
## # … with 1 more variable: year <int>

head( penguins[ ,sapply(penguins, is.double)] )

## # A tibble: 6 x 2
##   bill_length_mm bill_depth_mm
##            <dbl>         <dbl>
## 1         39.1          18.7  
## 2         39.5          17.400
## 3         40.300        18    
## 4         NA            NA    
## 5         36.7          19.3  
## 6         39.300        20.6

48 / 58

Challenge

Can you make return the opposite? In other words - all those that are not double?

02:00

49 / 58

head( penguins[ ,!sapply(penguins, is.double)] )

## # A tibble: 6 x 6
##   species island    flipper_length_mm body_mass_g sex     year
##   <fct>   <fct>                 <int>       <int> <fct>  <int>
## 1 Adelie  Torgersen               181        3750 male    2007
## 2 Adelie  Torgersen               186        3800 female  2007
## 3 Adelie  Torgersen               195        3250 female  2007
## 4 Adelie  Torgersen                NA          NA <NA>    2007
## 5 Adelie  Torgersen               193        3450 female  2007
## 6 Adelie  Torgersen               190        3650 male    2007

50 / 58

`vapply`

As you can probably see, simplifying can be really helpful for interactive work.

51 / 58

`vapply`

As you can probably see, simplifying can be really helpful for interactive work.

BUT

51 / 58

`vapply`

As you can probably see, simplifying can be really helpful for interactive work.

BUT

Not ideal for programmatic work - need to be able to reliably predict the output

51 / 58

`vapply`

As you can probably see, simplifying can be really helpful for interactive work.

BUT

Not ideal for programmatic work - need to be able to reliably predict the output
vapply solves this issue.

51 / 58

vapply(mtcars, mean, FUN.VALUE = double(1))

##        mpg        cyl       disp         hp       drat         wt       qsec         vs 
##  20.090625   6.187500 230.721875 146.687500   3.596563   3.217250  17.848750   0.437500 
##         am       gear       carb 
##   0.406250   3.687500   2.812500

vapply(penguins, is.double, FUN.VALUE = character(1))

## Error in vapply(penguins, is.double, FUN.VALUE = character(1)): values must be type 'character',
##  but FUN(X[[1]]) result is type 'logical'

vapply(penguins, is.double, FUN.VALUE = logical(1))

##           species            island    bill_length_mm     bill_depth_mm 
##             FALSE             FALSE              TRUE              TRUE 
## flipper_length_mm       body_mass_g               sex              year 
##             FALSE             FALSE             FALSE             FALSE

52 / 58

Coercion with `vapply`

If it can coerce the vector without loss of information, it will

vapply(penguins, is.double, FUN.VALUE = double(1))

##           species            island    bill_length_mm     bill_depth_mm 
##                 0                 0                 1                 1 
## flipper_length_mm       body_mass_g               sex              year 
##                 0                 0                 0                 0

53 / 58

Count missing data

vapply(airquality, function(col) {
  sum(is.na(col))
  },
  double(1)
)

##   Ozone Solar.R    Wind    Temp   Month     Day 
##      37       7       0       0       0       0

54 / 58

`sapply` alternative

For interactive work, the code on the previous slide is maybe too much. Could be reduced to

sapply(airquality, function(col) sum(is.na(col)))

##   Ozone Solar.R    Wind    Temp   Month     Day 
##      37       7       0       0       0       0

55 / 58

Summary

for loops are incredibly flexible and there's nothing inherently "wrong" about them

56 / 58

Summary

for loops are incredibly flexible and there's nothing inherently "wrong" about them
- Do require more text, and often repetitive text, which can lead to errors/bugs

56 / 58

Summary

for loops are incredibly flexible and there's nothing inherently "wrong" about them
- Do require more text, and often repetitive text, which can lead to errors/bugs
- The flexibility can actually be more of a curse than a blessing

56 / 58

Summary

The lapply family of functions help put the focus on a given function, and what values are being looped through the function

57 / 58

Summary

The lapply family of functions help put the focus on a given function, and what values are being looped through the function
- lapply will always return a list

57 / 58

Summary

The lapply family of functions help put the focus on a given function, and what values are being looped through the function
- lapply will always return a list
- sapply will try to simplify, which is problematic for programming, but fine for interactive work

57 / 58

Summary

The lapply family of functions help put the focus on a given function, and what values are being looped through the function
- lapply will always return a list
- sapply will try to simplify, which is problematic for programming, but fine for interactive work
- vapply is strict, and will only return the type specified

57 / 58

Next time

Lab 1

58 / 58

↑, ←, 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

Intro to iteration with Base R

Daniel Anderson

Week 2, Class 1

Agenda

Agenda

Learning objectives

Basic overview: for loops

Basic overview: for loops

Basic overview: for loops

Basic overview: for loops

Another basic example

Simulate tossing a coin, record results

Another basic example

Simulate tossing a coin, record results

Another basic example

Simulate tossing a coin, record results

Growing vectors

Growing vectors

Growing vectors

Example

Result

You try

Answer

Quick style note

Quick style note

Running the loop

Running the loop

Another example

Another example

Another example

Write for loop

List/data frame

tidyverse

Base R Method

Skipping iterations

Skipping iterations

Breaking loops

*apply

lapply

lapply

Revisiting our simulation with n=10n=10

The lapply version

Some more examples

Loop through a data frame

Some more examples

Loop through a data frame

Add a condition

Add a second condition

Passing arguments

Simulation again

Mimic dplyr::group_by

Your turn

Produce separate plots

Your turn

Saving

Saving

Saving

Saving w/for loop

Save plots to an object (list)

Saving w/for loop

Save plots to an object (list)

Specify file names/directory

Saving

You try!

Variants of lapply

Variants of lapply

Examples

Our simulation

Challenge

vapply

vapply

vapply

vapply

Coercion with vapply

Count missing data

sapply alternative

Summary

Summary

Summary

Summary

Basic overview: `for` loops

Basic overview: `for` loops

Basic overview: `for` loops

Basic overview: `for` loops

Write `for` loop

`lapply`

`lapply`

Revisiting our simulation with $n = 10$

The `lapply` version

Mimic `dplyr::group_by`

Saving w/`for` loop

Saving w/`for` loop

Variants of `lapply`

Variants of `lapply`

`vapply`

`vapply`

`vapply`

`vapply`

Coercion with `vapply`

`sapply` alternative