Echo™ User Manual

Abstract

Echo™ is a MATLAB-based software package designed for robust and scalable analysis of complex data workflows. An alternative to tedious, error-prone conventional processes, Echo is based on three transformative principles for data analysis: self-describing data, name-based indexing, and dynamic resource allocation. The software takes an object-oriented approach to data analysis, intimately connecting measurement data with associated metadata. Echo operations in an analysis workflow automatically track and merge metadata and computation parameters to provide a complete history of the process used to generate final results, while automated figure and report generation tools eliminate the potential to mislabel those results. History reporting and visualization methods provide straightforward auditability of analysis processes. Furthermore, name-based indexing on metadata greatly improves code readability for analyst collaboration and reduces opportunities for errors to occur. Echo efficiently manages large data sets using a framework that seamlessly allocates resources such that only the necessary computations to produce a given result are executed. Echo provides a versatile and extensible framework, allowing advanced users to add their own tools and data classes tailored to their own specific needs. Applying these transformative principles and powerful features, Echo greatly improves analyst efficiency and quality of results in many application areas.