rstudioconf.Rmd

---
params:
  lesson: "rstudio::conf2020"
  title: "Super tips, super lessons, and super powers from `R` super users"
  bookchapter_name: "Link to RStudio conference talks and workshop slides"    
  bookchapter_section: "https://github.com/EmilHvitfeldt/RStudioConf2020Slides"    
  functions: ""
  packages: ""      
  # end inputs ---------------------------------------------------------------
header-includes: \usepackage{float}
always_allow_html: yes
output:
  html_document:
    code_folding: show
---
  
```{r, setup, echo = FALSE, cache = FALSE, include = FALSE}
options(width=100)
knitr::opts_chunk$set(
  eval = T, # run all code
  echo = TRUE, # show code chunks in output 
  tidy = TRUE, # make output as tidy
  message = FALSE,  # mask all messages
  warning = FALSE, # mask all warnings 
  comment = "",
  tidy.opts=list(width.cutoff=100), # set width of code chunks in output
  size="small" # set code chunk size
  )
```
\

<!-- install packages -->
```{r, load packages, eval=T, include=T, cache=F, message=F, warning=F, results='hide',echo=F}
packages <- c("ggplot2","ggthemes","dplyr","tidyverse","zoo","RColorBrewer","viridis","plyr")   
if (require(packages)) {
    install.packages(packages,dependencies = T)
    require(packages)
    # load tvthemes
    devtools::install_github("Ryo-N7/tvthemes")
}
lapply(packages,library,character.only=T)  
```

<!-- ____________________________________________________________________________ -->
<!-- ____________________________________________________________________________ -->
<!-- ____________________________________________________________________________ -->
<!-- start body -->

# `r paste0(params$lesson,": ",params$title)`    
\  

Functions for `r params$lesson`  
`r params$functions`    
\    

Packages for `r params$lesson`          
`r params$packages`        
\    

# Agenda 

Notes from the in San Francisco on Jan 29-30, 2020 

Things to add  
- http://www.rebeccabarter.com/blog/2020-02-05_rstudio_conf/    
- joe cheng css and bootstrap  
- ryan tiempe using multiple packages 
- dewey and ggplot 
- tina cormier and spatial data? 
- pins package  
- `xgb` package (training/ML)  
- [Link to rconf recorded talks](https://resources.rstudio.com/rstudio-conf-2020)      
  
Weren't able to attend the conference? Fear not. The `R` community is sharing their talk and workshop slides. Thanks to [Emil Hvitfeldt](https://www.hvitfeldt.me/blog/) for collating everything.  

### [`r params$bookchapter_name`](`r params$bookchapter_section`). 
\  

<!-- ----------------------- image --------------------------- -->
<div align="center">
  <img src="img/rstudio.png" style=width:50%>
</div>
<!-- ----------------------- image --------------------------- -->
\    


<!--  end yaml template------------------------------------------------------- -->  

# Super useful and futuristic `R` tips and tricks  

Below are some useful tips and tricks I gleaned from presentations at the conference. It's a mix of tidyverse titbits for making workflow easier and future `R` development geared toward applied things like user experience and interactive apps, which is where `R`, science communication, and data analysis/viz is already deeply entrenched.

These examples were presented by super `R` users at the conference. I had to share some to show you how to bend `R` to your will to do equal parts good and evil. Links to presenters included if you want further info.   

## Pre installation notes   
Some packages may require installation from remote sources, such as Github, to which you'll need to prepare your comp with a software package downloader. Below is the quick and nasty way and more details can be found [here](
https://rstats.wtf/set-up-an-r-dev-environment.html#windows-system-prep
).  

For Mac, the software package compiler is called **Xcode Command Line Tools**. First, download it using the Terminal/command line:  
```{shell}
xcode-select --install
```

Then check you have everything set up in `R` by running the below command in `R`. You should then see the below confirmation. 
```{r, echo=T, results='markup'}
devtools::has_devel()
```
Your system should now be ready to download open source, in-development packages and package binaries.    

May run into this issue for Mac OSX Mojave: https://forums.developer.apple.com/thread/104296#317543

## Plotting 

Use user-defined variables in your `ggplot`   

- dewey and ggplot 

This throws an error because `ggplot` can't recognise the character string.

```{r,eval=F}
require(ggplot2)
my_theme <- theme_classic()
colour_var <- "class"
facet_var <- "drv"

ggplot(mpg) +
  geom_point(aes(displ, hwy, colour = colour_var)) +
  facet_wrap(vars(facet_var)) + my_theme
```


Placing `.data` in front of your variables and wrapping them with double square braces '[[]]' solves this.  
```{r}
require(ggplot2)
colour_var <- "class"
facet_var <- "drv"

ggplot(mpg) +
  geom_point(aes(displ, hwy, colour = .data[[colour_var]])) +
  facet_wrap(vars(.data[[facet_var]])) + my_theme
```

# More advanced stuff  

Integrating css and bootstrap