The R way
Note to BB: remember to start recording.
As we've seen R is powerful, flexible, and wide-ranging.
It's also sprawling, disorganized, and use-at-your-own risk.
Base packages, widely-used packages and packages by frequent R contributors are more trustworthy.
Some of the 'features' of R
- R is inconsistent in its naming conventions
- Some functions are 'camelCase'; others are 'dot.separated'; others 'use_underscores'. For programmers used to OOP, the use of dots is maddening.
- Multiple OOP paradigms and OOP is not as clean or full-featured as other languages
- R has multiple graphics packages that different functions use for default plot construction (
base
, grid
, lattice
, and ggplot2
)
- R has multiple packages and functions to do the same analysis
- R obscures some of what happens behind the scenes in a way that can be confusing and lead to errors: type coercion/casting, recycling, etc.
Any other things that you found frustrating and don't think are well-designed?
R tutorials and materials
Some of this material was borrowed/modified from Jared Knowles' bootcamp
CRAN has manuals and contributed learning material.
D-Lab holds workshops on R and other topics, as well as having drop-in office hours.
Some good books