BjornFJohansson / pydna

Clone with Python! Data structures for double stranded DNA & simulation of homologous recombination, Gibson assembly, cut & paste cloning.
Other
166 stars 45 forks source link

Motivations for pydna #39

Open BjornFJohansson opened 7 years ago

BjornFJohansson commented 7 years ago

Jupyter Notebooks and reproducible data science https://markwoodbridge.com/2017/03/05/jupyter-reproducible-science.html https://groups.google.com/forum/#!topic/jupyter/6pQIarRmrsc


https://towardsdatascience.com/5-tools-for-reproducible-data-science-c099c6b881e5

Reproducibility supports collaboration. Rarely does a data science project take place in isolation. In most situations, data scientists work together with other data scientists and with other teams to see a project through to integration into a business process. In order to collaborate effectively, it is important that other people can repeat, build on and maintain your work.

Reproducibility also supports efficiency. To be able to work most efficiently it is essential that you and your colleagues can build on the work that you produce. If results or processes cannot be accurately repeated then it is very difficult to develop on top of existing work and instead, you will find that you have to start a project all over again.

Reproducibility builds trust. As previously stated, data science is a discipline built on probability and experimentation. In this field trust in results is extremely important in order to develop buy-in for projects and to work effectively with other teams.