leftout_conditional_statements.Rmd

---
title: Conditional statements
---

<br>
<div class = "blue">
## Learning objectives

* Understand how to manipulate and transform data
* Be able to start with a single use and build up to iteration
* Use for loops, apply functions, and purrr to iterate
* Be able to write functions to cleanly iterate code
</div>

<br>


## Conditional Statements

Sometimes, you want to do something with your code, but only if a certain condition is true. There are a couple main ways to do this.

### `if` and `else`

You can use combinations of `if` and `else` to create conditional statements. Here's a quick example:

```{r if_loop}
for (i in 1:10) {
  if (i < 5) {
    print(paste(i, "is less than 5"))
  } else {
    print(paste(i, "is greater than or equal to 5"))
  }
}
```

Here we've combined a couple techniques: we've used a for loop to go through a sequence of values, and for each value we've printed a statement based on a condition that our value meets.

### `case_when`

Sometimes you might want to do a bunch conditional statements together, but typing out a ton of nested if-else statements can be unwieldy and prone to typos. A really useful function is the tidyverse's `case_when`. You feed it a dataframe and then use a series of two-sided formulas where the left-hand side determines which values you want, and the right supplies the result. Here's an example where we take the `mtcars` dataframe and add a column called `car_size`. If the car's weight is over 3.5 **or** it has 8 cylinders, we call it "big". If it's not big, but its weight is over 2.5, then it's medium. If neither of these conditions is met (denoted by `TRUE`), then we call it "small".

```{r}
mtcars %>% 
  mutate(
    car_size = case_when(
    wt > 3.5 | cyl == 8 ~ "big",
    wt > 2.5 ~ "medium",
    TRUE ~ "small"
    )
  )
```

<div class = "blue">
### Challenge

Using the `iris` data frame (this is built in to R), create a new variable that categorizes petal length into three groups:  

* small (less than or equal to the 1st quartile)
* medium (between the 1st and 3rd quartiles) 
* large (greater than or equal to the 3rd quartile)  

Explore the iris data using `summary(iris$Petal.Length)`, to see the petal length distribution.

Then use your function of choice: `ifelse()` or `case_when()` to make a new variable named `petal.length.cat` based on the conditions listed above.

<details>
<summary>ANSWER</summary>
```{r challenge1, purl=FALSE}

iris$petal.length.cat <- ifelse(iris$Petal.Length <= 1.6, "low", 
                          ifelse(iris$Petal.Length > 1.6 & 
                                   iris$Petal.Length < 5.1, "medium",
                           "high"))

iris %>% 
  mutate(
    petal.length.cat = case_when(
      Petal.Length <= 1.6 ~ "small",
      Petal.Length > 1.6 & Petal.Length < 5.1 ~ "medium",
      TRUE ~ "large"
    )
  )


```
</details>
</div>

<br>
### Conditional `map` 

Let's throw it back to the `map` family for a sec. Sometimes you might only want to map a function to part of a dataframe. `map_if` allows you to give the data, a condition for the data to meet, and the function you want to apply to the data that meet the condition. Here, we'll map `as.character` to the columns of the iris dataset that meet the condition `is.factor`.

```{r}
str(iris)

iris %>%
  map_if(is.factor, as.character) %>%
  str()
```

`map_at` does something similar, but it allows you to directly specify the locations you'd like to map the function to, using either names or positions.

```{r}
mtcars %>% 
  map_at(c("cyl", "am"), as.character) %>% 
  str()
```