Projekt "data.table.r"

This service agreement concerns the R library "data.table", with the aim of modernising its implementation in three milestones: 1) modularize the core methods in data.table, 2) introduce modern parallel processing, and 3) update the fast CSV reader for better performance, and to handle malformed data in a way controllable and visible to the user. Achieving these goals entails studying the current implementation to map the logical relationships between its subparts and devise a better scheme as per the state of the art; researching the best methods to implement parallel processing for the library’s functionality, taking into consideration the nature of processing it performs and the technical capabilities and limitations of the programming language and execution environment; and finally, improve performance in critical functions, and design and implement functionality for improved logging, error handling and reporting. data.table is an improved implementation of a core data structure in the R programming language that is used by virtually every component built on the language.