Coursera R Mentoring

Mentoring records and my helping articles on Coursera (Last update: July 29, 2019)

Some questions have been raised frequently in the Coursera R Programming course forum. Here I organized those questions and collected my replies to them.

Please note that these articles are not intended to be self-sufficient resources for learning R. If you are interested in R language, here are some good starts:

Data Science Specialization of JHU on Coursera (free auditing)
swirl package of R (R and swirl installation required)
R語言翻轉教室 (only for Mandarin-speaking users)

Besides, RStudio and R users around the world have contributed many cheatsheets which can be referred to anytime. Check RStudio Cheat Sheets for them.

For advanced R users. If you are interested in more detailed mechanisms of R grammar, you may refer to Advanced R written by Hadley Wickham, and Advanced R Programming.

BTW, for those who would like to take courses from Coursera but cannot afford the cost, I have some suggestions (written in Mandarin).

Getting started

Why R but not Excel: Here I listed some advantages of R over Excel from my point of view.
Use R or RStudio: RStudio can’t be executed without R. Still they have different advantages.
Chinese user names cause errors: If your computer user name comprise Chinese characters and you get errors when running RStudio, please refer to this article.
Use <- or =: Using either <- or = to assign values in R is a question related to personal preferences. I shared my own opinions here.
Why strptime returns NAs: Languages from different system locales may cause some abnormalities. However we can change the locale settings within R scope without changing the whole OS.

Data types

“Vectors” in R: Vectors in R don’t have mathematical sense. Yet “mathematical vectors” and matrix operations can still be performed.
Number types: R has class “integer” and “numeric”, they are different in memory management and capability to be manipulated.

Data preprocessing

Difference between [[]] and $: There are various extract operators in R. In this short paragragh I explain the difference between [[]] and $, and provide a link to an even more comprehensive article written by another Community Mentor.
Factor levels and labels: Categorical data are called “factor” in R. This article explains what are factor levels and factor labels respectively.
Split data: Data can be splited with specified categories. However the length of the factor variables matters.
Paste strings: Pasting various strings together is a plausible way to create variables. Here I introduce the two critical arguments, sep and collapse, of paste function in R.
Argument MARGIN for apply: apply allows users to execute a function over dimensions of matrices or arrays. What if we assign multiple dimensions at a time?

Comparison and ranking of data

Difference between == and %in%: Both == and %in% are often used to find matches in data. Yet, they have one difference when we have multiple inputs to be compared in parallel with the target.
Order, rank, and sort: These three functions are commonly used to make data into orders, but they are confusing with one another. This article tries to offer clarifications.

Functional programming

Difference between print and return from a function: Both print and return make objects “visible” outside the executed function, but only return truly returns values that can be saved for later use. Examples are provided in this article.

Lexical scoping of R language

Lexical scoping: This article demonstrates what would happen if an object is defined not only outside but also inside the function.
Constructor functions: With knowledge of lexical scoping, we are able to write a function which creates other functions.
Optimizing parameters: Another example showing how lexical scoping can help you optimize your interested parameters.

Generic functions

How mean was created: The source codes of mean cannot be seen by simply typing mean in the console. This article demonstrates how to see them.