This repo contains the R code, blog post Rmd file, knit html file, and perhaps most interestingly, a link to launch an Jupyter binder from which you can spin-up an Rstudio session into this repo! From there you can run the code and play with it right in your browser. This cool feature is brought to you by the stellar folks who made the rocker/binder image and those at https://mybinder.org/. This is a really new feature, so hiccups are not unlikely.
This post is based on the hyperparameter grid search example, but I am going to use it as a platform to go over some of the cool features of purrr that make it possible to put such an analysis in this tibble format. Further, I hope this post gives people some examples that make the idea of purrr "click"; I know it took me some time to get there. By no means a primer on purrr, the text will hopefully make some connections between the ideas of list-columns, purrr::map() functions, and purrr:nest() to show off what I interpret as the Tidy-Purrr philosophy. The part about using future to parallelize this routine is presented towards the end of the post.
Hypurrr-ameter Grid Search with Purrr and Future
Click the button above (into a new tab) to launch this repo into an Rstudio binder in your browser! After clicking the button, wait a moment for the https://mybinder.org/ service to build this repo into a new binder. Once built, this will open a new cloud based Rstudio session right in your browser. From there, you can click in the Files to the RMD folder and then click on the purrr_gridsearch_blog.Rmd to open the blog post markdown. Also, there is a strealined version of the main code example at /CODE/purrr_grid_search_plot.R. Run, edit, modify the code and have fun!