When Should I Use Python or R?

Python and R are commonly used, versatile programming languages for data science and analytics. Unlike commercial tools such as SAS and SPSS, both languages are open-source, free for anyone to download. However, both have different strengths and weaknesses meaning that the language you use will depend on your specific use case. Ultimately, both languages can accomplish nearly any data science task to various extents, from data manipulation and automation to ad-hoc analysis and exploring datasets. Users may leverage both languages for different purposes, e.g., conducting early-stage data analysis and exploration in R, then switching to Python when it’s time to build a production API. If you are new to learning programming languages, it is best to stay focused on learning how to break down and think through a problem, rather than trying to become a R or Python expert right out of the gate. 

• People with no software engineering experience often find R easier to learn than Python.
• Statistical models can be written with only a few lines of code - very efficient!
• Functionalities can be written in several different ways - R is not known for consistent syntax.
• Users can easily process complex functions in R.

• People with software engineering experience often find Python easier to learn than R.
• Simple syntax makes coding and debugging easy.
• Indentation of code affects meaning.
• Any piece of functionality is always written the same way - consistent syntax.
• Python is flexible when you need to create something brand new. Developers can also use it for scripting websites or other applications.

• It is easier to develop baseline competence in R, but the intricacies of advanced functionalities make it more difficult to develop expertise broadly across R’s libraries.
• Because syntax and functionality will vary from library-to-library, users need to be well versed in researching R skills on-line (e.g., Google, Stack Overflow) to learn.

• Python’s focus on readability and simplicity makes its learning curve relatively linear and smooth.
• While the ability to research various functionalities online is important, the consistent syntax of Python makes learning new packages and modules a straightforward task.

• R has a rich ecosystem of cutting-edge interface packages available to communicate between open-source languages.
• R users can string workflows together, which is especially useful for data analysis.
• Packages are collections of R functions, data, and compiled code, and can be installed
 in R with one line.
  
• R documentation is often readily available for most libraries (typically includes example code).
  
• Complementary offerings, such as “RStudio”, provide R users with an easy-to-use interface for writing, editing, and deploying code.

• Python has a robust ecosystem, the programming syntax is simple, and commands mimic the English language.
• Python code is syntactically clear, elegant, easy to type, and easy to interpret.
• Python is great for building data science pipelines and machine learning products integrated with web frameworks at scale. However, be mindful of dependencies when installing Python libraries.
  
• The Python Package Index (PyPi) and Anaconda are repositories of Python software with all libraries. Users can contribute to these repositories, but it’s a bit complicated in practice to do so.

• R has many functionalities for data analysis.
• R is great for statistical analysis, which are often cited in academic journals.
• While built around a command line, many R users work inside of RStudio. This environment includes a data editor, debugging support, and a window to hold graphics.

• As a general-purpose programming language, Python is useful beyond just data analysis.
• Python is popular for its code readability, speed, and diverse functionalities.
• Python is great for mathematical computation and learning how algorithms work.
• Python has a high ease of deployment and reproducibility.

• R was developed by statisticians, not coders, meaning that it was not designed to
 make coding easier.
  
• R must be learned library-by-library – it has no common syntax.
• Researching packages for use in R can be time consuming.
  
• There are many dependencies between R libraries.
  
• Poorly written R code runs slowly.
  
• R is not as popular as Python for deep learning and NLP - though this may change
 over time.
  

• Python has fewer purpose-specific libraries for data science than R.
• Working in Python requires rigorous testing, as errors show up in runtime.
• Visualizations in Python are more convoluted and not as eye-pleasing or informative as in R.