2024-07-11 –, Method (1.5)
Strategic Data Handling in Julia: Embracing Base Types for Business Impact
This talk caters to data analysts and scientists transitioning to Julia who are used to Pandas-centered workflows. We dissect the pros and cons of employing DataFrames versus Julia’s base types within a business setting, providing use case insights from our industry applications.
The talk targets Julia coders, who may not fully exploit its performance
capabilities due to a reliance on DataFrames. We will discuss the balance between development time, robustness, and performance, emphasizing time-to-value. Our aim is to provide attendees with strategies to maximize business outcomes when resources are constrained.
Attendees will learn to craft performant Julia code by employing arrays, structs, matrices, and dictionaries, which not only boost performance but also simplify logic testing. We’ll offer practical examples and refactoring tactics to transition from DataFrames to Julia’s more potent base types. Furthermore, we'll provide insight into the decision-making process for choosing the right data structure in varied business scenarios.
Participants will leave with a solid grasp of when to opt for DataFrames or Julia’s base types, grounded in practicality for business use. They will appreciate Julia's prowess in type stability and memory management for superior data handling. Resources for continued education and best practices will be shared to aid attendees in becoming adept Julia programmers for commercial applications.
Christian Hower is a Senior Data Scientist at Lexis Nexis focused on using data science to drive commercial outcomes.