Leveraging the R Packages googlesheets4 and ggplot2 for Fabric Shopping

I’ve become quite comfortable with Google API packages in my current workplace, where I collaborate with enthusiastic and creative Google Suite users to integrate their data and projects with SIS data to create data tools and dashboards. This has inspired me to use googlesheets4 in a recent personal project, shopping for quilting fabrics.

About

In October of 2024, I took a quilting class after having started sewing earlier in the year. To take my sewing skills to the next level, the course requirements asked me to show up to class #1 with 5-9 woven cotton fabrics.

Color choice is the part of crafts I feel least comfortable with. Sewing, like my knitting, and like much of the programming and data analysis I do, can be quite formulaic. I feel comfortable troubleshooting straightforward problems with clear solutions. Color choice feels like more of a design choice, although I’d love to take color theory coursework to help it feel more structured, and so my pre-class homework felt overwhelming.

To overcome the overthinking of color choices, I pulled hex code from some fabric shop websites, kept them in a Google Sheets spreadsheet with their respective URLs, and tested what a quilt might look like using the hex codes and ggplot2.

This combination of crafts and R is very simple, but it was a joyful way to use some functions I work with daily for a very me purpose.

Code

Load Libraries

Here are the packages I used. ARTofR is constantly on rotation for me, as I love the beauty of a thoughtfully organized R file. It’s useless here in this .Rmd for my distill site, but I’ll still show it love.

googlesheets4 is massively important for my work in OUSD. I work for a department of frequent Google Drive users, and I’ve googlesheets4-ed my way to some beautiful integration with student information system data with program data managed in Drive.

#install and load librarian package if needed
if (!require(librarian)) {
  install.packages("librarian")
  library(librarian)
}

# load packages
shelf(tidyverse, 
      here, 
      ARTofR, 
      ggplot2, 
      gridExtra, 
      googlesheets4)

Get Hex Codes and Randomize Color Selection

As I shopped for fabric, when I noticed colors that inspired me in the required fabric specifications, I’d grab the URL and throw it in this Google Sheet. I’d also use a Google Chrome Extension for picking colors from the web, then add the hex code to the sheet.

When I copy-paste from the color picker Extension, the hashtag isn’t included, so I add that below and then define a random selection of 5-9 colors to play around with when visualizing my potential quilt.

#  As I've been fabric shopping, I've kept a Google Sheet with URLs for
#  fabrics of interest and hex codes for the main colors from those fabrics.
hex_codes_data <- read_sheet('https://docs.google.com/spreadsheets/d/1bCFgHr1KwrBOcmKiO_k-jonU9haHtGD3Orw6rOVbw00/edit?gid=0#gid=0')

# Add '#' to the hex codes
hex_codes_data <- hex_codes_data %>%
  mutate(hex_code = paste0('#', 
                           hex_code))

# Define the number of colors to randomly select (between 5 and 9)
n_colors <- sample(5:9, 1)

# Randomly sample the hex codes
custom_palette <- sample(hex_codes_data$hex_code, n_colors)

Create a Sample Quilt Block

Using those randomized colors, I applied a custom function below. It utilizes ggplot2 to produced a 3x3 grid of 9-patch quilt blocks, which are a basic America quilt block. This is not what my quilt class quilt would be, but it allowed me to get a sense of how my colors might look together as blocks.

I ran this many times, examined the outcomes, and saved the palettes that I liked as additional tabs in my Google Sheets document. This way, I could call them back later as input to my function when making final decisions about fabric choices.

# Function to create a quilt block with a checkered pattern of two colors
create_quilt_block <- function(custom_palette) {
  # Create a data frame for grid coordinates
  blocks <- expand.grid(x = 1:3, 
                        y = 1:3)
  
  # Select two random colors for the block
  selected_colors <- sample(c(custom_palette), 2)
  
  # Assign colors in a checkered pattern (alternating between the two colors)
  blocks$color <- ifelse((blocks$x + blocks$y) %% 2 == 0, 
                         selected_colors[1], 
                         selected_colors[2])
  
  # Create the plot
  ggplot(blocks, aes(x = x, 
                     y = y, 
                     fill = color)) +
    geom_tile(color = "white", size = 1) +  # Create tiles with white borders
    scale_fill_identity() +                 # Use the exact colors provided
    coord_equal() +                         # Ensure square tiles
    theme_void()                            # No axis or labels
}

# Function to create a quilt with multiple blocks
create_quilt <- function(colors, 
                         n_blocks = 9) {
  
  plots <- lapply(1:n_blocks, function(x) create_quilt_block(colors))
  grid.arrange(grobs = plots, ncol = 3)  # Arrange quilt blocks in a grid
  
}

# Generate a quilt with 9 blocks using the randomly selected color palette
create_quilt(custom_palette, n_blocks = 9)

I’m very inspired by the work of the Generative aRt community online, but due to my color-choice hesitancy, I feel timid to explore Generative aRt, myself. I think I feel similarly about physical aRt. This exercise was perhaps an attempt at Generative cRafts, which I think I’ve just made up :)

Quilt, IRL!

Finally, here’s a look at the results of this fabric-shopping tool:

Here are some sample blocks I made throughout the class. Can you spy the 9 patch block? Or variations on it?

In this photo, my friend, Halina, and I are showing off the finished quilt front, with sashing added between the blocks! I think the colors are great!