Trying package development

I have always wanted to try building a package. Serious people seem to build packages. At least, serious people seem to know how to build packages.

But what to make? I was thinking of packaging the GreatSchools scraper but I’m a bit nervous that it might be too difficult or annoying to maintain. This may still happen, but not today. I need something simpler.

Independently, I’ve been inspired to try making art with R. I’ll write more about this later, but I went to an interview with artist Shirley Wu about how she got started in data visualization. She said she wanted to draw realistic flowers. This seemed like a good place to start. I have much more to say about making art with R but that, too, is for another day.

So I had a go at both! I made a little script that uses dplyr::geom_curve to draw little flowers. They’re not particularly realistic (yet!) – they’re a bit hacky, too symmetric, they need to be filled – but it’s a start.

One of my favorite people on Twitter, Jesse Mostipak, did a Twitch stream on package development. What perfect timing! I followed along (badly!) but I did it!

So, a new seed for the garden! A digital floweR

#devtools::install_github("enricomanlapig/floweR")

library(floweR)
library(dplyr)
library(ggplot2)


num_petals <- 7

df <- data.frame(
  group = c(rep("A", num_petals),
            rep("B", num_petals),
            rep("C", num_petals)),
  metric = c(runif(num_petals, min = 1, max = 2)*0.5,
             runif(num_petals, min = 1, max = 2),
             runif(num_petals, min = 1, max = 2)*1.2)
)

df %>%
  draw_flowers(my_grouping_var = group, 
               my_metric_var = metric, 
               my_color_var = metric, 
               my_hole_size = 0.5, 
               my_curvature = 0.5, 
               my_angle = 100, 
               my_lwd = 7) +
  
  scale_colour_gradient(low = "purple", high = "pink") + 
  scale_fill_manual(values = c("violet", "orange", "yellow")) + 
  theme(legend.position = "none")

#ggsave("featured.png")