several of the R chunks are unnecessarily complex. there's base::colMeans, stats::kmeans, graphics::pairs, etc. R's got the rvest package now, which is comparable to BeutifulSoup. Having used both extensively for database related work, "saving to databases" is as easy in R as it is in Python, and maybe even easier with the ORM-like support in the dplyr package. Python's been playing catch-up in the data analysis sphere lately, but with every gain that Python makes, R is already going to be that much further ahead. However, I second the other commenter's remark that Python is ahead in the deep learning domain.
