Making Functions for Rtistry

# Making Functions for Rtistry
### Meghan Harris, MPH
### R Ladies Johannesburg
### Tuesday, May 24th, 2022 | (24/05/2022)

---

<h1><span>Reach Out</span></h1>

<div class = "contact">
<br>
<a href="https://twitter.com/meghansharris"><img class = "icon", src ="images/twitter.png">@meghansharris</a>
<br>
<a href="https://thetidytrekker.com"><img class = "icon", src = "images/web.png">TheTidyTrekker.com</a>
<br>
<a href="https://github.com/Meghansaha"><img class = "icon", src = "images/github.png">@meghansaha</a>
<br>
<a href="https://www.linkedin.com/in/meghan-harris/"><img class = "icon", src = "images/linkedin.png">meghan-harris</a>
<br>
<a href="mailto:meghansa@buffalo.edu"><img class = "icon", src= "images/email.png">meghansa@buffalo.edu</a>

</div>
</div>

---
class: light, animated, fadeIn

<h1><span>What Are We Doing Today?</span></h1>

- What Are Functions in R?

- Why Use Functions in Rtistry?

- Using Base R for Data

- Using Purrr for Data

- Types of Functions for Rtistry

- Live Coding Example

---
class: dark, animated, middle, fadeIn

# <center>What Are Functions in R?</center>

*<center>"Work Smart, Not Hard"</center>*

---
class: light, animated, fadeIn

<h1><span>What Are Functions in R?</span></h1>

Functions are EVERYWHERE in R and you've most likely have used some already even if you haven't made any yourself:
<br>
--

- Functions in R can be considered *objects* that we can use to **do something.**

- According to the book [Advanced R](https://adv-r.hadley.nz/functions.html#function-fundamentals) by [Hadley Wickham](https://twitter.com/hadleywickham), Most functions are made up of three parts:
  - `formals()` - arguments that control how the function's called
  - `body()` - the code inside of the function
  - `environment()` - the structure that determines how the function finds the values associated the names in the function

- In this presentation, we'll focus mainly on **`formals`** and the **`body`** components of custom functions that can be used in rtistry.

---
class: light, animated, fadeIn

<h1><span>What Are Functions in R?</span></h1>

Let's look at an example of a basic function:

```r
# Creating a function named "rtistry_message":
rtistry_message <- function(adjective){
  
  # Setting up our custom message:
  todays_message = paste("Today is a", adjective, "Day for Rtistry!")
  
  # Returning our custom message to the console:
  return(todays_message)
}
```

```r
# Using our function:
rtistry_message("Great")
```

```
## [1] "Today is a Great Day for Rtistry!"
```

The **`adjective`** argument is a `formal()` that will control how the function is called. **Everything** inside of the curly brackets `{ }` is the `body()` of the function. In rtistry, we can create simple or complex functions to assist in the creation of our pieces.

---

# <center>Why Use Functions in Rtistry?</center>

*<center>"No Really. Stop Making Things Hard"</center>*

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Build up a tool box of functionalities, especially for things you'd like to do often.

--
<br><br>
<center><img class = "special", src = "images/art set.jpg" width = "300px", height = "300px"></center>

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Streamline code and reduce any manual (copy and paste) work

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

<p align = "right"> <sub>Thank you <a href="https://twitter.com/StarTrek_Lt"> Zane(@StarTrek_Lt)</a><br>For reminding me <a href= "https://carbon.now.sh/">Carbon</a> exists<br> and for reviewing the slides 😊</sub></p>

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Streamline code and reduce any manual (copy and paste) work

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Streamline code and reduce any manual (copy and paste) work

```r
# Load in your libraries====
library(ggplot2)
library(dplyr)
library(purrr)

#Setting seed for reproducibility====
set.seed(1129)

# Set up some options====
x_transformer <- c(0,15,30)
rectangle_groups <- c("left","middle","right")
rectangle_colors <- c("#82954B", "#AC7D88", "#143F6B")
rectangle_borders <- map_chr(rectangle_colors, ~colorRampPalette(c(.x,"#000000"))(10)[5])

# Create a "base" rectangle====
rectangle <- tibble(x = c(seq(0,10, length.out = 100),
                               rep(10,100),
                               seq(10,0, length.out = 100),
                               rep(0,100)),
                         y = c(rep(0,100),
                               seq(0,20, length.out = 100), 
                               rep(20,100),
                               seq(20,0, length.out = 100)))
```

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Streamline code and reduce any manual (copy and paste) work

```r
# Use Purrr functions to apply options to the base rectangle and create the data needed====
rectangles_df <- pmap_df(list(x_transformer,
                             rectangle_groups,
                             rectangle_colors,
                             rectangle_borders), ~rectangle %>%
                                                  mutate(x = x + ..1,
                                                         group = ..2,
                                                         fill = ..3,
                                                         color = ..4))
# Plot the piece====
rectangles_df  %>%
  ggplot(aes(x = x, y = y, group = group)) +
  theme_void()+
  geom_polygon(fill = rectangles_df$fill , 
               color = rectangles_df$color, 
               size = 1.5, 
               position = position_jitter(width = .05, height = .02)) +
  theme(plot.background = element_rect(fill = "#F1DDBF"))
```

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Streamline code and reduce any manual (copy and paste) work

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Make it easier to create variations in your work

```r
# Load in your libraries====
library(ggplot2)
library(dplyr)
library(purrr)

#Setting seed for reproducibility====
set.seed(1129)

# Set up some options====
x_transformer <- c(0,15,30)
rectange_groups <- c("right","middle","left")
rectangle_colors <- c("#1C658C", "#83BD75", "#CDBE78")
back_color <- "#363062"

# Just wrap everything in a custom function for better readability and ease====
rectangle_maker <- function(transformer, groups, colors){...}

# Use the custom function to make the data====
rectangles_df <- rectangle_maker(x_transformer, rectange_groups, rectangle_colors)
```

???
Rectangle colors =  blue, green, yellow
<br>
background color = purple

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Make it easier to create variations in your work

```r
# Plot the piece====
rectangles_df  %>%
  ggplot(aes(x = x, y = y, group = group)) +
  theme_void()+
  geom_polygon(fill = rectangles_df$fill , 
               color = rectangles_df$color, 
               size = 1.5, 
               position = position_jitter(width = .05, height = .02)) +
  theme(plot.background = element_rect(fill = back_color))
```

???
Using previous rectangle ex - easily change the colors
<br>
Background color object is put directly into the ggplot's theme function, we could put this in the function if we place the ggplot function in the "rectangle_maker" function as well

---
class: light, animated, fadeIn

<h1><span>Why Use Functions in Rtistry?</span></h1>

- Make it easier to create variations in your work

---

# <center>Using Base R for Data</center>

*<center>"Just like Vanilla Ice Cream 🍦, Base R is underrated"</center>*

---
class: light, animated, fadeIn

<h1><span>Using Base R for Data</span></h1>

We've already seen some examples of how we can create functions in R with the `function()` directive. Just like with many processes in R, you can make functions completely using Base R without any extra packages, if you're an R purist.

--
- There's no limit to what functions you can use in Base R to make your data but here's a few that are in my toolbox:
<br>
<br>

--
    -  <a href="https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/function"> `function()`</a>
--
    -  <a href="https://www.rdocumentation.org/packages/memisc/versions/0.99.30.7/topics/Sapply"> `sapply()` and `lapply()`</a>
--
    -  <a href="https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/sample"> `sample()`</a>
--
    -  <a href="https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/rep"> `rep()`</a>
--
    - <a href="https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/expand.grid"> `expand.grid()`</a>
--
    - <a href="https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/Control">Control Flows</a>

???
Quick reference to some functions I use frequently, mentioned in First rtistry workshop, slides available in my GitHub repo for this talk
---

# <center>Using Purrr for Data</center>

*<center>"Sometimes I wonder why Purrr has three R's"</center>*

---
class: light, animated, fadeIn

<h1><span>Using Purrr for Data</span></h1>

<a href="https://purrr.tidyverse.org/">Purrr</a> is a Tidyverse package that <i>"enhances R's functional programming toolkit by providing a complete and consistent set of tools for working with functions and vectors."</i>

--
- The Purrr package has lot of useful functions that can help with your rtistry. Here's just a few of my most commonly used functions:
<br>
<br>

--
    -  <a href="https://purrr.tidyverse.org/reference/map.html"> The `map()` Family</a>
--
    -  <a href="https://www.rdocumentation.org/packages/memisc/versions/0.99.30.7/topics/Sapply"> The `map2()` and `pmap()` Families</a>
--
    -  <a href="https://purrr.tidyverse.org/reference/keep.html"> `keep()` and `discard()` Functions</a>
--
    -  <a href="https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/rep"> `modify()` Functions</a>

---
class: dark, animated, middle, fadeIn

# <center>Types of Functions in Rtistry</center>

*<center>"There's no rules in Rtistry. Do what you want..."</center>*

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

The longer you do rtistry you may realize that you may start to develop a certain style. You may find yourself needing to do repeated types of operations. Likewise, you may get to a point where you'll be looking for new solutions to new challenges. Your journey may lead you to creating the following types of basic functions:

--
-   Repetition

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

```r
#Function for repeating vertical bars===
library(tidyverse)
vertical_bars <- function(n_bars, color_pal, background_color){
  
  #Setting iterations for "n" amount of bars===
  iterations = 1:n_bars
  
  #Creating a color gradient based on supplied colors values and desired "n"===
  bar_colors = colorRampPalette(color_pal)(n_bars)
  
  #Creating a "base" bar to copy and repeat over===
  base_bar = tibble(x = c(0,1,1,0),
                    y = c(0,0,10,10),
                    group = 0)
  
  #Creating all the data needed to repeat the base bar and apply the generated color palette===
  bars <- map2_df(iterations,
                  bar_colors, ~base_bar %>%
                                 mutate(x = x + 1.5*.x,
                                        group = group + .x,
                                        color = .y))
  #Creating the actual ggplot===
  bars %>%
    ggplot(aes(x,y, group = group))+
    theme_void()+
    geom_polygon(fill = bars$color, color = "#000000", size = 1)+
    theme(plot.background = element_rect(fill = background_color))}
```
???
For the sake of time, I won't go into this, code is available on the slides on GitHub
<br>
But here for any of you that want to go back and reproduce this image
---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

```r
#Options for the Piece===
color_pal <- c("#363062", "#4D4C7D", "#827397", "#E9D5DA") # A purple-ish color palette
back_color <- "#383838" # Some type of dark gray

# Using the function
vertical_bars(10, color_pal, back_color)
```

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

-   Randomization

```r
grid_maker <- function(rows,cols){...}
```

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

-   Helpers/Plug-ins

```r
# Data from another function===
squiggle_mess <- horiz_lines(0,10,15,10)

# Helper function to plot the data and add variation to size and background color===
plot_helper <- function(wide_size, small_size, background_color){
  
# Printing the piece===
squiggle_mess  %>%
  ggplot(aes(x,y, group = group))+
  theme_void()+
  geom_path(size = wide_size, color = "#000000", 
            lineend = "round", alpha = .7)+
  geom_path(size = small_size, color = "#ffffff", 
            linetype = 2, position = position_jitter(height = .09))+
  theme(plot.background = element_rect(fill = background_color))+
  coord_cartesian(expand = TRUE, xlim = c(0,20))
}
```

???
helpers/plug-ins can be used to build off of existing functions/pieces to introduce variation easily
---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

-   Helpers/Plug-ins

```r
plot_helper(wide_size = 10, small_size = .1, background_color = "purple")
```

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

-   Helpers/Plug-ins

```r
plot_helper(wide_size = 20, small_size = 5, background_color = "green")
```

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

-   Helpers/Plug-ins

```r
plot_helper(wide_size = 2, small_size = 5, background_color = "orange")
```

---
class: light, animated, fadeIn

<h1><span>Types of Functions in Rtistry</span></h1>

-   Procedural/Algorithmic

```r
# A 200-ish line function that procedurally creates circles in a "packing" fashion===

circle_packer <- function(n,min_x,max_x,min_y,max_y){...}
```

???
getting more involved may lead to procedure/algorithmic when you want to start calculating things, implementing complex logic, etc.
<br>
still working on this so not publicly available 
---
class: dark, animated, middle, fadeIn

# <center>Live Coding Example</center>

*<center>"AKA: The part where Meghan probably forgets everything she knows because, pressure."</center>*