Have you ever wished that creating an app was as easy as calling a function? Now it is!

1. Write an R function,
2. pass it to appify,
3. specify the inputs.

Done. There is your app!

appify is meant to be used with OpenCPU. Include an Rmarkdown document or website in your R-package. Set the output directory in yaml to ../inst/www/. Inlcude an R script in your package's R directory with a command like this rmarkdown::render_site(input = "_app/") assuming you've put the files for your rmarkdown website in _app/. Templates will become available soon. Check the demo for a working example.

install.packages("opencpu")
install.packages("devtools")
devtools::install_github("retowyss/appify")

Have a look at the template package on Github retowyss/appify-demo, clone it, build it and run it with OpenCPU.

opencpu::ocpu_start_app("appify-demo")

Load the package.

require(appify)

Write your R function.

iris_clustering <- function(color_1, color_2, color_3) {
  require(ggplot2)
  clusters <- kmeans(iris[, 1:4], centers = 3, nstart = 3)
  
  iris_clustered <- cbind(iris, clusters = factor(clusters$cluster))
  
  ggplot(iris_clustered, aes(x = Sepal.Width, y = Petal.Width, color = clusters)) +
    geom_point() +
    scale_color_manual(values = c(color_1, color_2, color_3)) +
    ggtitle("Kmeans Clustering") + 
    theme_minimal()
}

appify has only one top level function appify. The function requires two inputs:

1. a function f and
2. a list inputs.

That's it.

appify(
  f = iris_clustering, 
  inputs = list(
    color_1 = dropdown_color("Color 1"),
    color_2 = dropdown_color("Color 2"),
    color_3 = dropdown_color("Color 3")
  )
)