As an R package, Shiny is ideal for building interactive web apps with interesting applications for a range of industry sectors from teaching and real estate to crime and life sciences.

This workshop will bring more insight into common thoughts people have about Shiny, such as, is it possible to develop more complex applications with Shiny? What would be the main limitations? Could R and Shiny be used as a general-purpose framework to develop web applications? What needs to happen to facilitate that? These are just some of the questions we will look to answer.

In the session we will explore the potential of web app development with Shiny, with particular focus on:
• Exploring the possibilities of Shiny
• Exploring different ventures within the Shiny framework
• Learn how to apply Shiny to app development

• Shiny functionality
• Creating a game loop
• Managing a global game state
• Interactivity as a useful feature

Plumber APIs are a great tool to integrate your R data science with the rest of your tech stack and to turn code into a data science product.

Interesting use-cases for plumber include:

• Display heart disease risk on a website using a Cox regression written in R

• Display the results of random forest model in a Power BI dashboard

• Integrate the impressive interactive and dynamic visualisations created in R in mobile phone applications

• Outsource statistical models to R within your Python software project

• Create a web scraping API to extract recent data from a website using rvest

In this session you will learn how easy it is to build APIs in R using the:

• Plumber package

• How to organise code for plumber APIs

• Using plumber-style comments to create API endpoints and filters

• Sending HTTP requests from multiple mediums; swagger, terminal, R and Postman

• Containerise plumber APIs for deployment

• Improving performance of plumber APIs

Background
• What is an API?
• What is plumber?
• An example of a plumber endpoint

Getting started exercises
• Setting up the plumber project as a package
• Using block annotations to specify endpoints
• Using serializers and parsers to change the request and response formats
• Sending HTTP GET requests

Intermediate exercises
• Sending HTTP POST requests
• Using global data to improve performance
• Understanding the request and response objects
• Adding an authorisation filter to authorise users

Advanced exercises
• Containerising the API with Docker
• Improving performance with future and promises

Explainable ML presents a significant opportunity to using data responsibly. Interesting use case include:

• Identifying mental health needs

• Distinguishing differentiation in job salaries

• Diagnosing skin diseases

In this session you will be able to:

• Describe explainable AI methods.

• Explain the importance of building explainable models to other stakeholders.

• Implement a range of XAI techniques.

• Introduction to Explainable AI / ML (XAI / XML)

• A practical example focussed on one R package

• Conclusions and next steps

Iteration (doing the same task multiple times) is a very common task in Data Science. A loop in R programming is an incredibly common way of achieving this, however they are often not the best solution. Instead, purrr (a package from the tidyverse) allows you to tackle iteration in a functional way, leading to cleaner and more readable code.

The purrr package offers a great suite of tools to allow you to elegantly solve numerous iteration challenges in R:

• Simultaneously fitting models to multiple groups

• Generating multiple reports with different input parameters

• A full understanding of the purrr package and how it can be used

• Amongst other things, we’ll show you how to fit over 140 statistical models and extract fit statistics for each of them in only 3 lines of code

• The ability to write code that is succinct, easy to read and reason about by colleagues

Iteration and the purrr package
• Iterating over each list element
• Applying custom functions
• Simplifying output

Working with lists
• Filtering list elements
• Joining lists

The wider map family
• Applying across multiple lists
• Side effects
• Using the index

Nested data
• Nested data frames
• Mutate and map
• Map for modelling & simulation
• Converting to nested data

If you have seen ambient temperature plots, electrocardiograms, or even stock market fluctuations, you have already seen a time series visualisation. Many prediction problems are dynamic and involve a time component.

How does time affect our perception of the data? Is the time scale important? Is the direction of time relevant? Sometimes cumulative effects are not visible with traditional statistical methods, because little increments stay under the radar. When a time component is present, it’s likely that the current state of our problem depends on the previous states. With time series visualisations we can capture changes that may otherwise go undetected.

In the session you will learn:

• The concept of time series

• The ability to demonstrate the limitations of ordinary plots

• How to visualise single and multiple time series

• How to visualise trends, seasonality, remainders

• Ability to perform and plot forecasts

Introduction

• What is a time series?
• What makes time series unique
• Time series tools in base R

Simple tasks
• Define time series objects
• Simulate your own time series
• Simple visualisations
• Missing data imputation

Intermediate tasks
• Time series decomposition
• Visualise trends, seasonality, remainders
• Looping over multiple time series
• Alternative visualisations
• Advanced tasks
• Visualising forecasts
• Visualising cross-validation
• Exploring additional R packages

Do I need a valid email address to buy a pass to EARL?

Yes, it is important you provide a valid email address when registering. We will be sending you the receipt for your purchase and other communications related to EARL 2022 at this address.

Can I buy multiple passes in one transaction?

Our ticketing system requires delegates to register and buy passes independently. You can however use EARL 4+1 Team Bundle options to buy a block of 5 passes in one transaction. If you want to buy a specific number of passes, please contact EARL HQ at earl@ascent.io and we can arrange this for you.

I am trying to purchase a ticket, but it is saying I am already registered for the event. What do I do?

Close your browser completely and open the EARL ticketing page again. Our system is based on unique emails so you may need to use another email address to register for the same event. If you want to buy a specific number of passes or you need any assistance, please contact EARL HQ at earl@ascent.io and we can arrange this for you.

I am part of the NHS/ an academic/ work in a non-profit. How do I access the discounted rate?

If you work in the NHS (National Health Service), non-profit sector or in a formal academic setting, you are eligible for discounted EARL passes. You can apply for a code via this form. Complete the form and once verified, we will send you your discount code. To book your EARL passes, select your tickets on the site as normal and enter your code in the box at the point of checkout.

The form requires you to provide proof of eligibility. We may at our discretion request additional proof of identity at the registration desk on the day(s) of the conference.

I have special dietary requirements, what should I do?

Please message us at earl@ascent.io to make us aware of any allergies or intolerances you may have. This will be communicated to the Tower Hotel to ensure your requirements are catered for.

I booked a ticket but might not be able to attend. What do I do?

If you contact us at earl@ascent.io we can organise a ticket transfer for you to another named delegate. For refund requests, please send an email to earl@ascent.io detailing the exceptional circumstances. Refunds will be dealt with on a case-by-case basis.

Do you provide support for accommodation bookings?

We do not book accommodation for our delegates however, the Tower Hotel is ideally situated within a short distance of hotel options. Please note, if you are attending the evening event at Drapers Hall, we will provide transport to the event from Tower Hotel, but you will need to organise your own transport from Drapers Hall back to your accommodation at the end of the evening.