While submitting calculations through the web is the simplest way to get started with Rowan, complex workflows can often benefit from being able to run and analyze calculations programmatically. Rowan provides API support for all calculations and workflows, making it easy to convert web-based workflows into high-performance scripts.
Many scientific simulation programs generate data through heterogenous or unstructured formats: energy values are often buried somewhere inside a large text file, or data is divided (or duplicated) between a series of different output files. This makes analysis difficult, time-consuming, and error-prone.
Rowan provides a high-level interface that's documented, consistent, and integrates easily with other parts of the scientific programming ecosystem. With just a few lines of code, you can integrate Rowan's entire suite of computations and workflows with the rest of your tools and modernize your chemical computation capabilities. Plus, Rowan's API automatically allocates resources through our cloud computing infrastructure, making it easy to scale workflows without having to manage high-performance computing in-house.
Rowan's API can be accessed through the open-source Python package rowan-python, which can be installed on any machine. This library allows users to submit, monitor, analyze, stop, and delete calculations through a single programming interface. Data is returned through the open-source stjames format, which can easily be converted into any number of formats for further processing.
Rowan's API can be run in "blocking" or "non-blocking" modes, allowing users to choose their desired level of abstraction for the task at hand. API calls are automatically linked to a user's account via secure API key: all jobs run through the API can be viewed in the web browser and vice versa, making it easy to troubleshoot and visualize calculations where necessary.
For more details and lots of example scripts, see the full API documentation.
