Posted by Ohene Akuoko
Hadley Wickhams’s talk was very enlightening to some of the old and the new functions of R. It was obvious listening throughtout of his hand in deveoping R, by his understanding of how to create and manuver through the “interactive environment”.
One thing that I enjoyed about his talk was some of his lamen defintions for some terms and concepts that may otherwise be difficult to grasp.For example, the Tidy data rules.. 1. All data in data frame 2. Make sure each value has a column. He also explained some tidy data tools prinples, like simply take one step at atime when using functions, Compose those steps in a pipe and and the newer concept of referenctial transparency. I also, learned that there are actually alot more impure functions than what I had assumed, based on his defintion of pure data.
The introduction of the pipe and its use was reinfoced in my mind as being a very effective tool to perform data analysis in R. As he mentioned the “pipe” was essentially designed for humans and not computers like basic R, without elimnating the option of basic R or any other functions. I also thought useful in his talk was concept of list-columns in tibble to create vectors out of single values, to be able to utilize polygons and map making which is something I would be interseted in using R for in the future.
Also looking into the future of R programming, it seems that gstat will be very useful and (I think) a gamechanger in the realm of producing visual statistics for scientist and data managers alike. The referential transparency tool, also seems like it will help develop more effective funtion creation and again would be something big in the future, though it is hard to grasps how it operates at the moment.