Investigate R (back-end) to Python (front-end)

[CharlotFranssen]

• or other front-end versions
• Create summary of findings

[WouterMulder9]

We can integrate R with Python using the package “rpy2”: https://marketsplash.com/tutorials/r/how-to-integrate-r-with-other-programming-languages/. However, instead of using Python, we can also use an R package instead called “shiny”: https://shiny.posit.co/. This package enables us to build a web application directly from R scripts. It is especially useful when the application primarily involves statistical analysis or data visualization.

Based on this cheatsheet: https://rstudio.github.io/cheatsheets/html/shiny.html, it looks quite easy to work with.

[WouterMulder9]

Maybe could add some extra information on the up- and downsides of both methods, which would not only include the perspective from the programmer, but also the ease of use for the user, especially considering that the users are not in posession of a very technical background. And thus should have a big impact on the eventual decision made

[QuintineSol]

Integration with rpy2 (Python Frontend): Pros (Programmer Perspective):

• Familiarity: If the development team is already proficient in Python, integrating an R backend using rpy2 allows them to leverage their existing skills and tools.
• Versatility: Python offers a wide range of libraries and frameworks for web development, providing flexibility in designing the frontend interface.
• Ecosystem: Python's ecosystem is vast and well-supported, offering extensive documentation and community support for problem-solving.

Cons (Programmer Perspective):

• Integration Complexity: Integrating R with Python using rpy2 adds an additional layer of complexity to the development process. Developers need to manage interoperability between two languages and handle potential compatibility issues.
• Performance Overhead: Transferring data between R and Python can introduce performance overhead, especially for large datasets or complex computations.
• Maintenance: Maintaining code written in two different languages may require additional effort and increase the risk of errors.

Ease of Use (End-User Perspective):

• Seamless Integration: End-users may not notice the backend/frontend distinction if the integration is seamless. However, any issues or inconsistencies in the interface could impact user experience negatively.
• Performance: If data processing or visualization is slow due to the integration overhead, it could affect the responsiveness of the application, leading to frustration for end-users.

Integration with Shiny (R Frontend): Pros (Programmer Perspective):

• Simplicity: Developing the frontend directly in R using the shiny package eliminates the need for integration with another language, simplifying the development process.
• Rapid Prototyping: shiny provides a high-level framework for building interactive web applications, allowing developers to quickly prototype and iterate on ideas.
• Tight Integration: Since both the backend and frontend are written in R, developers have direct access to R's powerful statistical and visualization capabilities, enabling seamless integration of analysis and visualization components.

Cons (Programmer Perspective):

• Limited Frontend Options: While shiny offers a convenient way to build web applications in R, it may lack the extensive frontend customization options available in other web development frameworks.
• Learning Curve: Developers unfamiliar with shiny may need to invest time in learning its specific syntax and conventions, potentially slowing down the development process initially.
• Scalability: While shiny is suitable for building small to medium-sized applications, it may not be the best choice for large-scale or complex projects due to performance limitations.

Ease of Use (End-User Perspective):

• Cohesive Experience: End-users interacting with a shiny application experience a cohesive interface where data analysis and visualization seamlessly integrate with user interactions, enhancing usability.
• Performance: Since both the frontend and backend are in R, there's typically minimal overhead in data transfer, resulting in faster and more responsive applications.
• Familiarity: Users familiar with R for data analysis may find shiny applications more intuitive to use, as they leverage similar syntax and conventions.