EdinbR: Thrifty analysis and testing
Title: Thrifty analysis and testing
Speakers: Mayuko Morgan and Nick Radcliffe
Date: Thursday 11th June 2026, 6.00PM - 7.00PM
Location: 2.55 The Bayes Centre The University of Edinburgh
Mayuko Morgan: The thrifty data chef’s guide to cook up insight on a shoestring
Python or R - it's not necessary to pick either, you can move between them. Both bring their own strengths, styles and a wealth of useful libraries. In this talk, I will walk you through a fun experiment I carried out for my master's dissertation: building a data mining and visualisation pipeline that leverages open data and open-source tools, and the best of both worlds of python and R (with an R-focus for my talk). I'll also show how a low-cost AI API can add a splash of extra flair!
Mayuko is currently immersed in her PhD research at Edinburgh Napier University, exploring the ares of personality-orientated dialogue systems for humanoid robots, following her completion of a Master's Degree in Data Engineering. Before starting her PhD, she embraced a squiggly, non-linear career trajectory that took her through diverse fields - from software development to environmental science and geographical information systems - though “data” has always been at the core of her passion!
Nick Radcliffe: Test-Driven Data Analysis: The Methodology, Library, CLI, and Book
Test-Driven Data Analysis (TDDA) is an approach to improving the quality of analytical processes and data. It is directly inspired by test-driven development (TDD) and overlaps with ideas in Reproroducible Research. This talk will describe the methodology and the TDDA Library, which has a CLI that can be used for data validation and test generation for any language. It will also outline Test-Driven Document Development (TDDD), which relates to but differs from computational documents.
Nick has worked in data science, machine learning, and operational research for about 40 years. He has worked in commercial data science, founding Quadstone (in 1995) and Stochastic Solutions (in 2007). He is also a Visiting Professor in Maths and Edinburgh University. He has been developing TDDA for about a decade, and published a book with that name in May 2026.