Michael Weidman, MathWorks
Mixed data (a combination of data types such as numbers, text, dates, categories, etc.) can be painful to organize and analyze. Whether it is “panel data” in finance, “experimental results” in the sciences, or the like, it has traditionally required the use of either multiple variables or generic data containers during analysis. The Dataset array in MATLAB’s Statistics and Machine Learning Toolbox offers a superior solution: a single, easy-to-use object that can import, analyze, and export the (possibly) mixed data.
In this webinar we use a case study of unbalanced panel data to highlight a typical workflow for and the features of Dataset arrays. We discuss features such as categorical arrays, name-based indexing, and joining datasets.
View MATLAB code from this webinar at MATLAB Central
About the Presenter: Michael Weidman earned a B.A. in physics from Harvard University and completed Part III of the Mathematical Tripos from DAMTP at the University of Cambridge. Michael joined MathWorks in June 2007, and he focuses on computational finance applications.
Recorded: 20 May 2009