All Times EDT
Keywords: Python, R, Notebooks, Functional Programming, Deep Learning
Python and R are traditions coming from neighboring lands, intermingling and trading ideas. Among each generation of citizens are new authors, who add their own embellishments, drop what they don’t need, and retain familiar anachronisms, puzzling new users but also paying homage. In this talk, we contrast how each expose the Deep Learning frameworks. We chronicle the rise of tidy data frames, the use of notebooks, and functional programming, from which a unique data science workflow has emerged. We also discuss the flaws of notebooks and the flexibility of RStudio. We illustrate each as idiomatic shells on top of backends like Stan, Spark, and Tensorflow, which give power to both R and Python, yet are harnessed differently by each community. Ultimately, we tend to use R for data analysis and visualization and Python for deep learning, yet mostly we ‘Do as the Romans Do’, and may adopt Julia and Swift in the future.