DSC: Dynamic Statistical Comparisons

Overview of DSC

DSC provides a framework for managing computational benchmarking experiments that compare several competing methods for a task across datasets or simulation scenarios.

DSC helps execute such comparisons in an organized and reproducible way, and provides convenient ways to query the results.

DSC is designed to help make these comparisons dynamic — that is, it is easy to extend by adding new methods or simulation scenarios. Hence the name, Dynamic Statistical Comparisons.

DSC is able to run benchmarks written in most languages that can be compiled or run as executables. DSC is particularly well-suited to run methods implemented in R and Python, the two predominant interactive programming languages in scientific research. DSC also provides support for experiments that combine R and Python.

DSC is implemented in Python 3, building on the SoS framework, with additional tools developed for R.

Getting started with DSC

If you are new to DSC, we recommend taking a look at the introductory DSC tutorial. This tutorial gives an overview of DSC's main features, and illustrates how DSC can be used to quickly implement a simple computational experiment.

If you would like to try DSC for yourself, follow the installation instructions to download and set up DSC on your computer. Then return to the introductory tutorial and try implementing the example on your computer.

If you would like to use DSC for your project, start by reading through the Introduction to DSC Syntax Part I and Part II. To learn more, we provide additional tutorials on other DSC topics, a DSC reference manual, and extended examples illustrating how DSC can be used to implement a variety of computational experiments.

If you have any questions or want to share some information with the developer / user community, please open a github issue.

Acknowledgement

This work is supported by the the Gordon and Betty Moore Foundation via an Investigator Award to Matthew Stephens, Grant GBMF4559, as part of the Data-Driven Discovery program.


Copyright (c) 2016-2018 Gao Wang, Matthew Stephens, Peter Carbonetto and contributors.
    Except where otherwise noted, this website is licensed under a Creative Commons Attribution 3.0 Unported License.