SATRDAY

The afrimapr project will create software building blocks in R and learning resources to facilitate the use of spatial data in health (and other) applications in Africa. The project will promote these resources to initiate a community of users and developers to maintain and improve them.

Co-authors:

• Dr Andy South (Liverpool School of Tropical Medicine, UK)
• Dr Paula Moraga, (Lancaster University, UK)
• Dr Julie-Anne Tangena (Liverpool School of Tropical Medicine, UK)
• Dr Robin Lovelace (University of Leeds, UK)
• Dr Margareth Gfrerer (Education Strategy Center, Ethiopia)

Our final outputs from R will often take on the form of a graph, map, infographic or other relevant visual. These visualisations are not devoid of creative elements. They require colour, and sometimes they require a multitude of custom palettes for the task at hand. How best can these be derived?

Do you wish you could drag and drop variables onto X and Y axes and easily map variables? Are you in a hurry (or just plain lazy) to type code by hand? The {esquisse} add-in lets you explore and visualise your data with {ggplot2} interactively and provides the generated code to reproduce the chart!

Advances in our understanding of the human genome have largely been based on European populations. More recently, a number of studies have been investigating African populations, which have a higher degree of genetic diversity. What better way to map them out than with a bit of help from sf?

Have you ever wanted a set of productivity boosting tools to help make your workflow more efficient? Do you ever wonder what anyone does with their .Rprofile? If you answered yes to both these questions then this lightning talk is for you.

My talk focuses on the construction of a generalized additive model (GAM) to relate healthcare service efficiency and operational metrics to clinic profitability (profit per day). My modelling used vtreat for R (John Mount and Nina Zumel) as it made the modelling process and output a tidy-treat!

Interpretability is still regarded as one of the main barriers of machine learning. How do we explain a model decision to a client? How do we measure the underlying interactions in a black box model? Let's have a look at Shapley Values as a model agnostic technique for interpreting ML models.

There is so much more to purrr than map(). purrr is a veritable treasure trove of functional awesomeness that can help transform the code you write and what it can do. In this talk, I illustrate what lies beyond map(), how to use it, and why it may be worth exploring purrr's hidden depths.

One of the most challenging tasks in data science is moving code from a development through to production. Build on top of Apache Spark, Databricks provides streamlined workflows and interactive workspaces that enable collaboration between Data scientists, data engineers and business analysts.

This talk shows the endless possibilities of R and shiny's capabilities in a production environment using docker and ShinyProxy. In this talk I will demonstrate how to build and deploy a shiny application (statistical ad hoc report) for each user in an isolated docker container on a linux server.

The idea that R is not a good fit for engineering production software is a catchphrase we hear a lot in the world of data science, and in the software engineering community in general. But the truth is that the R community has been creating tools for building robust software for years—tools that have been proven to be reliable and efficient, making R a legitimate language for production.

With all the available tooling, the real challenge of sending R to production is not technical—it's all a matter of attitude. In other words, the real challenge is our ability to adopt a software engineering mindset when building tools with R.

So where do we start our journey to production quality with R? What are these tools that help us building reliable products? How do we think about R with 'deploy to production' in mind? These are the questions, amongst others, that Colin will try to answer during his keynote, which will hopefully help you become more confident about sending R into production.

Resurrection plants are capable of surviving total water loss and prolonged metabolic inactivity in the dry state - plant biltong for years! Once water is available, they miraculously recover within hours. How R they capable of this? Come find out! ;)

According to UN Africa is affected by Culture,Traditions and beliefs in the journey towards achieving SDG 5. There are legal frameworks that have been put In place to ensure that countries get here. This study uses R to conduct a NEP analysis that will show which countries have reached the threshold

In the past few years, the R Community has been making conscious efforts to widen the participation of underrepresented groups. This talk will review some of the initiatives that have contributed to this effort, including Forwards, a task force set up by the R Foundation.

A lot of progress has been made and the R Community is now recognized for its inclusiveness. However, there is still room for improvement, particularly in the technical side of the R project (developer meetings, package development, R core development) and in the general user community for specific groups (people living in underserved regions or people with disabilities that affect access). This talk will consider these challenges and discuss some ways we might move forwards to achieve greater equity and inclusion.

Visiting a new place on holiday or travelling for work? What can you see from where you’re staying? What are the shortest routes? To answer these questions, you’d probably ordinarily reach for your phone and open up Google Maps. However, as a data scientist, with a specific set constraints and conditions (including having my running shoes on), I want to delve a bit deeper, explore and map out my options. My first tool to reach for is R. Then, the Open Source Routing Machine (OSRM).

OSRM is a high-performance routing engine for calculating the shortest paths through a road network. These calculations are available via Google Maps. However, queries against a local OSRM server are orders of magnitude faster than Google Maps (and cheaper!). The osrm package for R exposes these routing data to a wide range of potential applications. In this talk I'll show how to easily spin up and provision an OSRM server and use it to solve some interesting spatial optimisation problems in R.

We've been gathering retail pricing data. As an exercise we took a look at how prices change around Valentine's Day and found some interesting things.

I will use Google Trend data which is like a hidden closet of thoughts, together with Prophet, an open-source library published by Facebook to predict trends in chocolate. 

A novel demonstration of the Funk FM styled factorization machine popularized by Simon Funk in 2006 (Netflix prize). My intent started off as a recommendation system but evolved into a way to impute pixels into corrupted images from a synthesized matrix.

As a Pythonista, moving to R for a new job was not the most natural thing. I had to learn so many things! 6 months into R, I would like to share a few things about:

• How Python differs from R
• Which R tools to learn as a beginner Rtist
• Which online resources can help ease the transition

At the time of this publication, the SARB still has no easy way of accessing their data through an programmatic interface (API). The package aims to fill this role by curating the data automatically from the SARB, updating a dedicated database and giving access through a simple function.

This talk is born out of the observed exponential infection growth of the Corona virus outbreak in China - which was not observed with MERS or SARS. With the underlying data and equipped with R - lets visualise the outbreak, run some simulations and understand the outbreak a little better.