Background
Scientific software workflows often rely on a mishmash of codebases, including low-level libraries, domain-specific open-source software, and self-developed and/or inherited toolboxes (whose original developer may or may not be around to pass on undocumented wisdom). This leads to proliferation of tools where people spend time reinventing wheels of variable quality, which jeopardizes the ideal of code being "re-runnable, repeatable, reproducible, reusable, and replicable". Beyond these requirements, low-level programming abstractions are often a distraction that get in the way of understanding the actual science. We developed Sciris as a means of streamlining the development of scientific software by making it easier to perform common tasks, similar to how PyTorch simplifies/extends Tensorflow, or how seaborn simplifies/extends Matplotlib.

Methods
Sciris "stands on the shoulders of giants", and as such is not intended as a replacement of other libraries, but rather as an interface that facilitates a more rapid and effective development process. Sciris provides classes and functions that simplify access to frequently used low-level functionality in the core libraries of the scientific Python ecosystem (such as NumPy and Matplotlib), as well as in libraries of broader scope (such as multiprocess for parallelization and pickle for saving and loading objects). Some of Sciris' key features include: ensuring consistent dictionary, list, and array types (e.g., enabling users to provide inputs as either lists or arrays); enabling ordered dictionary elements to be accessed by index as well as key; simplifying datetime arithmetic by allowing date input in multiple formats, including strings; simplifying the saving and loading of files and complex objects; and simplifying the parallel execution of code. Some of Sciris' key design features include: (a) brevity through simple interfaces (such as plot3d() to directly create a 3D plot), (b) "dejargonification" (such as findnearest() to find the element in an array nearest to the target value), (c) flexible exception handling (allowing users to quickly switch between being strict or forgiving), and (d) version management (such as saving figures with the Git branch/commit information in the metadata). Sciris also forms the basis of ScirisWeb, an additional set of tools for building Flask-based Python webapps.

Conclusion
During the COVID-19 pandemic, Sciris proved invaluable for allowing us to rapidly develop Covasim, a Python-based modeling tool that was used for research and policy decisions in dozens of countries . Sciris also forms the backbone of numerous other scientific software libraries, such as Optima HIV and HIPtool. Sciris makes writing scientific code in Python faster, more pleasant, and more accessible, especially for people without extensive training in software development.