No More Loopy Code: Data Science Goes Functional

Data and their analysis play a key role in our modern society. Deriving correct results is crucial to ensure reliable decision-making, leading to better outcomes in business, healthcare, governance and daily life. Yet, in data science we frequently encounter a makeshift approach, where various tools are tied together with untested procedural glue code. This talk explores the benefits of applying functional programming principles in data science. Emphasising immutability, pure functions and the benefits of strong type systems, we demonstrate how these concepts lead to more reliable and maintainable data workflows. Practical examples will showcase improved code clarity, verifiable specifications, parallel processing capabilities and simplified debugging. Learn how integrating functional programming techniques can transform your data science practices, resulting in more predictable and reproducible outcomes. This talk is aimed both at data scientists and everybody with an interest in analysing data in a reliable and robust manner. Prior knowledge of functional programming is helpful, but not required.