On June 24, Daniel Bankmann introduced an audience consisting of approximately 10 interested coworkers and students to Zenodo and Conda.

The talk adressed an issue that researchers may face on the regular: Trying to validate or reproduce results of a research article not written by yourself. Often, you are not able to do so, because the supplementary software or the research code is either decades old or has changed over time, leaving no possibility to obtain the older version that is required to validate results. This issue raises a question: How may we enable other researchers to reproduce results – or, in other words, how to make our own research reproducible?

A possible solution to this problem is Conda, a tool for managing and deploying applications, environments and packages. Conda is able to switch between different environments that may contain different versions of programs. Furthermore, it allows you to build a package that contains exactly the version of your code and complementary software that a colleague might need to run to validate your results. Daniel gave an example on how to create a package for a simple toolbox written in Python. Furthermore, Daniel introduced the web service Zenodo, which enables you to link in publications to the exact state of research software that was used for generating results or examples.

Tools Seminar: Reproducible Research II – Zenodo and Conda