This is a repository template for the re:Clojure / SciCloj workshop "Wrangling Arrays with dtype-next", given in association with the 2021 re:Clojure conference, and developed by David Sletten and Ethan Miller

This workshop will introduce dtype-next, explain its position in the Clojure data science ecosystem, and introduce the key concepts and techniques necessary for working with its performant buffers/arrays.

This is a github template repository. That means you can use this repository to create a new repository of your own in order to play around with dtype-next further. See here for a guide on how to create a repository from this template.

What is in this repository?

• A deps.edn file containing the dependencies you need to work with dtype-next. Really this is just one dependence: the tech.ml.dataset library -- that provides a column-based dataset for data analysis -- and which includes the latest stable version of dtype-next.

• A namespace for starting your own exploration that already includes typical require statements for dtype-next, i.e. src/playground/main.clj.

• The code that we will cover in our workshop, see the src/workshop/*.

1. Create a buffer containing a fibonacci sequence using the make-reader function.

2. Try to use the argfilter and indexed-buffer functions to filter a numerical buffer by a < or > operation on the existing values.

Try to complete some of these exercises for Numpy using dtype-next: link