Apply Family of Functions

class: center, middle, inverse, title-slide

.title[
# Apply Family of Functions
]
.author[
### Justin Post
]

---

layout: true

---

# Efficient Code

For loops vs Vectorized Functions

---

# `apply()` Family

- `apply()` family of functions *pretty* fast

- Check `help(apply)`!

+ We'll look at `apply()`, `sapply()`, `lapply()`

---

# `apply()` Family

- `apply()` family of functions *pretty* fast

- Check `help(apply)`!

+ We'll look at `apply()`, `sapply()`, `lapply()`

- Consider our Batting data set

``` r
library(Lahman)
my_batting <- Batting[, c("playerID", "teamID", "G", "AB", "R", "H", "X2B", "X3B", "HR")] |>
  as_tibble()
my_batting
```

```
## # A tibble: 128,598 × 9
##   playerID  teamID     G    AB     R     H   X2B   X3B    HR
##   <chr>     <fct>  <int> <int> <int> <int> <int> <int> <int>
## 1 aardsda01 SFN       11     0     0     0     0     0     0
## 2 aardsda01 CHN       45     2     0     0     0     0     0
## 3 aardsda01 CHA       25     0     0     0     0     0     0
## 4 aardsda01 BOS       47     1     0     0     0     0     0
## 5 aardsda01 SEA       73     0     0     0     0     0     0
## # ℹ 128,593 more rows
```

---

# `apply()` Family

- Use `apply()` to find summary for the batting data

``` r
apply(X = my_batting, 
      MARGIN = 2, 
      FUN = summary,
      na.rm = TRUE)
```

```
##           playerID teamID      G     AB      R      H    X2B    X3B     HR
## Length      128598 128598 128598 128598 128598 128598 128598 128598 128598
## N.unique     24011    256    165    699    167    250     64     32     69
## N.blank          0      0      0      0      0      0      0      0      0
## Min.nchar        5      2      3      3      3      3      2      2      2
## Max.nchar        9      3      3      3      3      3      2      2      2
```

---

# `apply()` Family

- Let's try it with just numeric data!

``` r
batting_summary <- apply(X = my_batting |>
                           select(where(is.numeric)), 
                         MARGIN = 2, 
                         FUN = summary,
                         na.rm = TRUE)
batting_summary
```

```
##                G       AB         R         H       X2B       X3B      HR
## Min.      1.0000   0.0000   0.00000   0.00000  0.000000  0.000000  0.0000
## 1st Qu.  11.0000   3.0000   0.00000   0.00000  0.000000  0.000000  0.0000
## Median   31.0000  40.0000   3.00000   7.00000  1.000000  0.000000  0.0000
## Mean     47.2958 129.3894  17.29095  33.76627  5.768674  1.154497  2.6883
## 3rd Qu.  71.0000 195.0000  24.00000  48.00000  8.000000  1.000000  2.0000
## Max.    165.0000 716.0000 198.00000 262.00000 67.000000 36.000000 73.0000
```

---

# Anonymous Functions

- We often use our own custom functions with the `apply()` family

+ Called anonymous functions or lambda functions

---

# Anonymous Functions

- We often use our own custom functions with the `apply()` family

+ Called anonymous functions or lambda functions

``` r
custom_batting_summary <- apply(X = my_batting |>
                                  select(where(is.numeric)), 
                                MARGIN = 2,
                                FUN = function(x){
                                  temp <- c(mean(x), sd(x))
                                  names(temp) <- c("mean", "sd")
                                  temp
                                }
)
custom_batting_summary
```

```
##             G       AB        R        H      X2B      X3B       HR
## mean 47.29580 129.3894 17.29095 33.76627 5.768674 1.154497 2.688300
## sd   45.83771 177.0342 26.95612 50.05455 9.272108 2.461973 6.197295
```

---

# Anonymous Functions

- Anonymous functions can take other arguments

``` r
custom_batting_summary <- apply(X = my_batting |>
                                  select(where(is.numeric)), 
                                MARGIN = 2,
                                FUN = function(x, trim){
                                  temp <- c(mean(x, trim), sd(x))
                                  names(temp) <- c("mean", "sd")
                                  temp
                                },
                                trim = 0.1
)
custom_batting_summary
```

```
##             G        AB        R        H      X2B      X3B       HR
## mean 40.69582  94.06721 11.15623 22.94881 3.630074 0.532465 1.045636
## sd   45.83771 177.03418 26.95612 50.05455 9.272108 2.461973 6.197295
```
---

# `lapply()`

- Use `lapply()` to apply function to lists

- Obtain a list object

``` r
set.seed(10)
my_list <- list(rnorm(100), runif(10), rgamma(40, shape = 1, rate = 1))
```

---

# `lapply()`

- Apply `mean()` function to each list element

``` r
lapply(X = my_list, FUN = mean)
```

```
## [[1]]
## [1] -0.1365489
## 
## [[2]]
## [1] 0.5997619
## 
## [[3]]
## [1] 1.108209
```

---

# `lapply()`

- To give additional arguments to `FUN` we add them on afterward

``` r
lapply(X = my_list, FUN = mean, trim = 0.1, na.rm = TRUE)
```

```
## [[1]]
## [1] -0.1359629
## 
## [[2]]
## [1] 0.6062252
## 
## [[3]]
## [1] 0.9563087
```

---

# `sapply()`

- Similar function but it attempts to simplify when possible

``` r
sapply(X = my_list, FUN = mean, trim = 0.1, na.rm = TRUE)
```

```
## [1] -0.1359629  0.6062252  0.9563087
```

---

# Recap!

- Vectorized functions fast!

- `apply()` family is sort of vectorized

- `lapply()` and `sapply()` to apply a function to a list

- `aggregate()`, `replicate()`, `tapply()` `vapply()`, and `mapply()` also exist!