For. Loop. Fun.

Objective

learn base commands to call key components of your data (nrow, ncol, $, [,],c(), etc.)
use base syntax to explore the logic behind forloops

Using a mock dataset ‘ToothGrowth’

Mock data in baseR

R has plenty of dataframes and matrices to play with, take a look below!

```{r lets look at data} data() # wow! these are all mock datasets in base R!

`ToothGrowth` contains data on the effects of vitamin C on tooth growth in 60 Guinea pigs. Each animal received one of three dose levels of vitamin C (0.5, 1, and 2 mg/day) by one of two delivery methods, orange juice or ascorbic acid (vitamin C as `VC`).

## Before getting started..

Lets assign `DF_tooth` as our working version of ToothGrowth in our exercises

```{r}
DF_tooth <- ToothGrowth # lets call it something else

Add some base commands to your tool belt!

Explore the data a bit - while we’re at it walk through some core commands

View() - open the whole dataframe in a window

{r view} View(DF_tooth) # opens a window with the data

head - shows you the first rows of your dataframe

```{r the first few} head(DF_tooth) # head shows your the first rows of the dataframe
```
```{r the first two}
head(DF_tooth, 2) # tell it how many rows you want to see
```
tail - shows you the last rows of your dataframe

```{r the last of ‘em’} tail(DF_tooth) # tail shows you the last rows of the dataframe
```
```{r the last one}

tail(DF_tooth, 1) # tell it how many rows you want to see
```

nrow - count the number of rows

nrow(DF_tooth) # prints a single integer for your object

ncol - count the number of columns

```{r how many columns are there}

ncol(DF_tooth) # prints a single integer for your object ```
dim - dimensions of your object
```
dim(DF_tooth) # 60 rows and 3 columns!
```
[,] - use brackets to call the “x and y” coordinates of your object, being your [row, column]

```{r call it out}

tail(DF_tooth, 1) # look at the last row again, we see the row number is 60 and there are three columns

last row is 60 and the first column is len - below == 23!

DF_tooth[60,1] # 23

last row is 60 and the second column is supp

DF_tooth[60,2] # ‘OJ’

last row is 60 and the third column is dose

DF_tooth[60,3] # 2

```
c() - concatenate, used to extract rows and columns or list values!

```{r combine}

c(1:5) # with integers!

c(‘one’, ‘of’, ‘me’, ‘is’, ‘order’, ‘out’, ‘of’) # with strings! ```

Activity: Print the first five rows of the ‘len’ column in DF_tooth
use the commands above to solve

```{r try it out}

head(DF_tooth) # we just want to print 4.2, 11.5, 7.3, 5.8, 6.4

try it below!

```

Loops, what are they good for? .. absolutely somethin’!

About for-loops iterate over a collection of objects, such as a vector, a list, a matrix, or a dataframe, and apply the same set of operations on each item of a given data structure.

Why join the for-ce?

keep your code clean, remove unnecessary repetitive code blocks
- streamline you QC, statistics, sanity checks, and more!
honest data carpentry - start with raw data to form something new & reproducible!
- create new columns based on unique criteria
- comment as you go, keep a bread crumb trail from old to new data files
(my opinion) keeps coding logic fresh!

How to do it

for to open your for loop
() set your identifier and call collection of objects
{} open and close - write what you want to apply to each object!

for ( in ) {

apply what is written here to each object

} close the loop

Example #1 (custom column): Next time floss the pigs

Turns out the students fed the Guinea pigs orange juice (supp = ‘OJ’) with extra pulp! It was on sale at Aldi and they couldn’t beat the bargin.

Unfortunately, OJ accumulated on the teeth and artificially inflated length measurements by 10%.

Let’s use a for loop to make a new len column len_new

for to open your for loop
() set your identifier and call collection of objects
- (i in 1:nrow(DF_tooth)) - our collection of objects are row-wise, meaning that the for loop with iterate commands for each row i of DF_tooth
{} - inserted below is an if else conditional statement to call row-wise boolean value before either calculating the new value when supplemented OJ or simply outputting the same length value when supplemented vitamin C.

```{r large pulp teeth, error = FALSE}

for (i in 1:nrow(DF_tooth)) { # open for loop, for i equal to 1:60

# do the following for when i is equal to 1, 2, 3, ..60

# this is one possible solution using a if else conditional statement!

if (DF_tooth[i,2] == ‘OJ’) { # if row i in the second column == ‘OJ’

DF_tooth$len_new[i] <- DF_tooth[i,1] + DF_tooth[i,1]*(0.1) # new row 'len_new' calculated

} else { # FALSE, second column == ‘VC’

DF_tooth$len_new[i] <- DF_tooth[i,1] # 'len_new' row is the same as len

} # end if statment

} # close for loop

View(DF_tooth) # look at the data!

## There's an ~~app~~ package for that!

-   `dplyr` has a nifty one liner that does the exact purpose as the for loop above!

-   input a conditional statement and the output if TRUE or FALSE using the command `if_else`

    -   `if_else(condition, TRUE, FALSE)`

```{r large pulp teeth, error = FALSE}
if ("dplyr" %in% rownames(installed.packages()) == 'FALSE') install.packages('dplyr')

library(dplyr) # load the library

?if_else # review the command page

# input the criteria - same as the for loop above

DF_tooth$len_new2 <- if_else(DF_tooth$supp == 'OJ',
                            (DF_tooth$len + DF_tooth$len*(0.1)), # TRUE
                            (DF_tooth$len)) # FALSE

View(DF_tooth)

Example #2 (plot loop): Flower power

Looping through data columns is an efficient method to plot large data sets, especially if you have custom plotting code! Below is a simple example outputing plots.

for to open your for loop
() set your identifier and call collection of objects
- (i in 1:(ncol(iris)-1)) - our collection of objects are column-wise, the for loop will iterate commands for each column i of iris except for the last column
{} - inserted below is an ggplot geom_boxplot for each column with numeric data

```{r plotting flower, error = FALSE} if (“ggplot2” %in% rownames(installed.packages()) == ‘FALSE’) install.packages(‘ggplot2’)

library(ggplot2) # load the library

iris # take a look at the dataset!

dim(iris) # 150 rows and 5 columns

head(iris) # first four rows are numeric data and last row is categorical ‘Species’

below we loop thorugh each iris numeric column and plot by species

for (i in 1:(ncol(iris)-1)) { # open for loop, for i equal to 1 - 4

loop_measurement_name <- colnames(iris[i]) # string to name the plot and y axis

plot <- ggplot(data = iris, aes(x=Species, y=iris[,i]), fill=Species) +
  geom_boxplot() +
  geom_jitter() +
  ylab(loop_measurement_name) + # name the y axis w
  ggtitle(loop_measurement_name) + # name the plot
  theme_classic()

print(plot)

} # close for loop

clock below to see the four unique boxplots!

-   Note: as opposed to `print()` we could use the unique identifier 'loop_measurement_name' to name and output pdfs or jpegs onto your computer!

## Activity! *For, if, else, case_when, Oh My!*

**About**: Tackle the puzzle below using (1) raw base R syntax using `for` with `if` and `else` conditional statements and (2) the dplyr alternative

A greenhouse experiment was originally planned to breed hybrid strains from each of the three iris species. However, the off-brand fertilizer you used caused each species to grow a trait unknown to iris ..ologists(?)!! Paying homage to the man-eating plant from Little Shop of Horrors, you name this discovery `Audrey_III`.

The allometric relationship of `Audrey_III.Length` with Pedal.Length is different for each species

-   setosa - `Audrey_III.Length`= Pedal.Length\^2.4

-   versicolor: `Audrey_III.Length`= Pedal.Length\^4.0

-   virginica: `Audrey_III.Length`= Pedal.Length\^3.0

**Objective**: create a new column `Audrey_III.Length` based on the relationship with Pedal.Length for each iris species

```{r for loop}

for () { # open for loop

      if () {        # condition

                     # if TRUE

      ) else if () { # next condition

                     # if TRUE for else if only

      } else {  

                     # if FALSE for all conditions

      } # end if statment

} # close  for loop

View(iris) # look at the data!

Objective: create a new column Audrey_III.Length based on the relationship with Pedal.Length for each iris species

```{r case_when} ?case_when

iris$Audrey_III.Length_NOloop <- # use case_when to take it!

View(iris) # when you finish, compare your Audrey_III results

```

For. Loop. Fun.

Using a mock dataset ‘ToothGrowth’

Mock data in baseR

Add some base commands to your tool belt!

last row is 60 and the first column is len - below == 23!

last row is 60 and the second column is supp

last row is 60 and the third column is dose

try it below!

Loops, what are they good for? .. absolutely somethin’!

Example #1 (custom column): Next time floss the pigs

Example #2 (plot loop): Flower power

below we loop thorugh each iris numeric column and plot by species

clock below to see the four unique boxplots!