tips_and_tricks.Rmd

---
title: Tips and Tricks
author: Robert McDonald
date: "`r format(Sys.Date(), '%B %d, %Y')`"
output:
  html_document:
    toc: true
---

```{r setup, echo=FALSE, message=FALSE}
library(tidyverse)
library(knitr)
opts_chunk$set(error = TRUE
               ,collapse = TRUE)
```



# Tips

## Setting up your working environment

* Make sure that you always start R with a clean environment
  * In RStudio:
    * In "Global Options", ***uncheck*** the box reading "Restore
    .RData into workspace at startup"
    * Set "Save workspace to .RData on exit" to ***Never***
  * In other editors, always answer "n" to the `Save workspace image?
    [y/n/c]` prompt when you exit R

* At least once, you should execute `options()` at the command line
  and take the time to look at the output. This shows you various
  options you can set globally for the current session.  Here are two
  you might wish to change to suit your preferences:
  * Output width: Set to 120 with `options(width=120)` (default is `r#
    getOption("width")`)
  * Number of digits: Set to 3 with `options(digits=3)` (default is `r#
    getOption("digits")`)
  * Set both at once: `options(width = 100, digits = 3)`
  * You can save and restore a set of options:
  ```{r}
  options('width', 'digits')
  saveopt <- options('width', 'digits')  ## save options
  saveopt
  options(width = 150, digits = 12)
  options('width', 'digits')
  options(saveopt)  ## restore 
  options('width', 'digits')
  ```
  
  Note that when you restart R, options return to their original defaults.

## Debugging

* First and most important: **read the error messages**. Even if you
  find them confusing, they convey information about what went wrong
  and where it occurred
* If you can't quickly find the problem, save your documents and
  restart R. Sometimes the problem goes away with a clear environment
* 

# Make sure you know which function you're using

* Function names are not unique. The same function name can be used in
  multiple packages, and the functions may do very different things.
  For example, `dplyr` uses `filter`, which is also provided by the
  `stats` package, and `summarize` is also provided by the `Hmisc`
  package. If you get an incomprehensible error message associated
  with a `dplyr` function name, one of the first things to try is to
  qualify the function name. Here is an example in which the first
  calculation fails but the second succeeds.^[Note that standard
  practice in Python is to qualify function names so that this
  confusion does not occur. For example you would write `import Pandas
  as pd`, and then call Pandas functions with a `pd.` prefix.]
  
  Note that if you use the British spelling,  `summarise`, this issue
  does not occur. (The `dplyr` package recognizes both spellings, the
  `Hmisc` package does not.)
  
```{r, message=FALSE}
library(Hmisc) ## some other package may have invoked this

mtcars %>% group_by(gear) %>% 
    summarize(N = n())

mtcars %>% group_by(gear) %>% 
    dplyr::summarize(N = n())

mtcars %>% group_by(gear) %>% 
    summarise(N = n())

```